Well, maybe if the tutorials out there spent less time being condescending and more time actively presenting the real paradigms instead of flimsy confusing stuff, it wouldn't be a problem.
"What's the right mouse button for?"
wrong: "it's a context-sensitive menu enabling access to control commands"
wrong: "it's like a scrapbook in which your least used situational commands are gathered and presented for your use"
right: "it's where your less common controls go. there're even rarer ones in the big menus. it works on pretty much everything. just try it out a lot; as long as you don't pick any menu items, nothing's gonna change, and you won't hurt anything."
1) Give them a simple straightforward explanation of what it does without jargon or metaphor
2) Encourage them to familiarize themselves with the control, being careful to note when such experimentation is inappropriate, even when it's never inappropriate
Not so hard. Out of curiosity, I sat through a biug chunk of the tutorial shipped with my new commodity PC; there were some things I didn't understand, and I wrote software for a living.
Perhaps hire fewer multimedia visionaries and more teachers next time you guys write intros.:D
Sorry, just saying luminaries wanted it banished and then not explaining left a bad taste in my mouth. I should have provided this link to a Herb Sutter article called "Can't Afford Export" (regarding getting rid of the export keyword) in the original post.
This isn't the only such article, but it's the only one I'm finding on a brief search; IIRC, I think I've also seen similar stances from Meyers and Dewhurst. Not certain, though, 'cause I can't seem to find them (yay 30 second searches.)
Comeau is the only 100% compiler; most everyone else hoses export, which is of arguable utility (and which many luminaries want to banish anyway.) For all practical purposes, Borland and Gnu have been ISO compliant for years. Intel's compiler is a monster. VC7 is actually pretty good, though VC6/eVC3 were just a mess. Sun's compiler is pretty strong. The Metaware and Gold Hills toolchains are very strong. ARM's ADS is very compliant (moreso than SDT was.) I don't honestly know the status of IBM's compilers, but I expect the best from them; same goes for SGI since their years as Cray.
Frankly, the chances that anyone has real use for anything modern compilers regularly get wrong is fairly slim. You must be *powerful* with the force to really scare G++.
That's a fairly mundane observation. I could just as easily suggest that "the fact is that if the language permits you to write anything you want, most people will write bad code."
In fact, that's quite what I'm suggesting is going on. If you have a high power toolset, you have a high powered way to write shitty code. No amount of language alteration will change that bad programmers write bad software. Just as it's not the STL's fault if you can't handle iterators, it's not the language's fault if you can't handle ranges. Don't blame your inability to write secure code on the language. Read a fucking book.
Any language that allows for someone making a living pointing out everything one shouldn't do needs more than a face-lift.
It's not exactly clear to me whether I'm taking a joke on as if it were serious. That said, this is a genuine sentiment with many, so I feel my rebuttal to be germane to them if not nessecarily to you.
The problem with language is that its domain is inherently difficult. The more expressive a language is, the deeper and more powerful structures it can express. Conversely, because making a language more expressive involves adding new methods of control, the more expressive a language is, the more difficult it is. You need to know how to generate the syntax which wields the new expression mediums, etc.
The problem as I see it is that the professional field requires a significantly higher degree of skill in expression than the amateur field wants to put forth. At first this seems like a tautology, but think about it: it's fairly rare that an experienced c++ programmer suggests there needs to be a starting-over. Given that few other dominant-in-their-day languages can make that statement, I think it's worth considering that maybe C++ is as horrible as it is by necessity.
Bjarne suggests that there are more than a dozen languages named D (which is stupid; c++ *is* D, as well as P; don't try to extends jokes you don't understand, folks, kthxbye) and another half dozen named E or P. That some of them are backwards compatible and *still* haven't takjen over should be an indication that C++ is doing at least something right, despite hordes of popular, well supported contenders.
I too hate a lot of things about C++. But frankly, even though it's not even close to my first language, I can't think of another language I like more for a generic problem domain. Sure, when it's web apps it's PHP; when it's database toys or mostly UI apps it's Delphi; et cetera. Still, when it's difficult, it's c++, every single time.
Any profession that allows for a whole subprofessions making a living pointing out what one should not do has become satisfyingly deep. Things which one can just jump into and be the best at are shallow. Art, writing, engineering, mathematics, induction, politics, philosophy. Each of these has books, tomes on what you should *not* do in order to be effective. In some fields this study of counterexamples has become so important that it elevates to its own subfield of study (fallacies and converses;) arguably that's happening in CS with AntiPatterns right now.
What I think you're seeing is a language which is way ahead of the game in community support. ;)
strncpy() has existed for quite some time, thanks. How about eliminating the coding practices that lead to buffer overflows? Oh, right, it's the *language's* fault. God forbid a language which is designed explicitly to allow you to do exactly what you want to in the nearest-to-machine way reasonable allow you to write security holes too.
Maybe instead of asking your toolset to design for you - security is a design process, not a patch process, which is why it works for qmail but not for sendmail - you should just learn to write secure code.
that Scott Meyers wrote 3 large volumes on the subject.
To bolster the point, there are a fair number of other authors which have written such series, and some of them arguably more difficult than the effective trio. Notably, Dewhurst's C++ Gotchas and Sutter's Exceptional/More Exceptional C++ are similar vein, similar-or-greater difficulty issues. None quite so well written as Meyers', though.
but this would effectively be a massive distributed DOS attack on spammers.
versus
In other words, you could host your Viagra-peddling site with a company that has a stringent no-spam policy, but a DNS lookup will point to a home user's compromised machine.
Attacking a spammer's resources only increases the spammer's impetus to steal resources. The further you push them underground, the harder they are to uproot when you get a real tool.
1) Few movies are double platter DVDs. You should be starting with 4.7g.
2) The original poster said DVD-quality. Transcoding to MPeg4 or DivX dramatically reduces size.
VCDs were thriving almost a decade ago. Compression/decompression technologies have since improved significantly, as has end-user processing power. A FoaF regularly downloads movies over broadband; takes him 15-20 min each. It takes me fifteen minutes just to get to a rental place, much less park, come back, deal with traffic, deal with rental and late fees, etc.
And while you're right that the trip to Blockbuster doesn't take ten hours, waiting for that high-school schmuck to scan two people's tapes does.
> How long would a typical DVD quality movie take to > download?
Less time than it takes to go to Blockbuster. Less gas, too.
Why no compulsory licenses?
on
Why Only Music?
·
· Score: 4, Interesting
Very simply put, because I don't want to pay for other people's right to steal, I don't want to underwrite the industry's estimates of what their failing business model, i mean, technology is costing them, and I don't want to be told that this is the alternative to allowing a puny entertainment industry to infringe upon my rights.
It's their problem, and it doesn't come out of my pocket, amortized or not, period. This is not a socialism. Stop pretending.
1000s of streams of content, some indistinguis...
on
TV's Tipping Point
·
· Score: 1
http://reinvent.the.whe.el/
The TV of the future you're describing is called teh intarweb, lady. Just so you know.
ICANN shouldn't have to sue anyone over a technical aspect of the Internet. They should have the tools to simply tell Verisign to do it and have it done quickly.
They are not suing. They are, in fact, leveraging their contract - their tool - and telling verisign to get it done and have it done quickly. Specifically, 36 hours. The thing about the business world is that if they didn't make sure that they were on strong grounds, if they demanded the service be taken down and then got sued, then they'd be indemnable for whatever money verisign made up that they lost on absent sitefinder service.
ICANN is doing the right thing, in fact the very thing that we're angry that VeriSign didn't do: they're checking that their actions are correct before undertaking them. ICANN has a responsibility to be proper and careful, rather than just running around swinging its arms like a bully (which some would say that it has done in the past.)
Look, you can't please everybody: if you do it fast people will say you didn't plan, and if you plan people will say you didn't do it fast enough. Don't you think it best that they do this in the way that's most difficult for VeriSign to prevent?
It's difficult to be the good guy.
And they should also have the means to simply cut Verisign out of the loop
As has been pointed out, they have implied that they will do just that in about 36 hours if their demands aren't met. As other/.ers have pointed out, they can just instruct the root servers to route around the damage.
(Of course, nobody seems to be pointing out that there's going to be the demand for some tremendous bandwidth and heavy servers pretty on-the-spot if they choose to do that. I find myself wondering which company will attempt to step up to the bat and steal the gold ring, if VeriSign fucks this up.)
There are actually a number of life-critical operating systems in play, a few of which have been certified by the NSA. DEC and IBM make very large zeros on this sort of thing from airports, et cetera.
...no one gets killed when Dr. Watson pops up and you have to restart Word.
Notably, lawsuits can be filed for things that just cost tremendous amounts of money. Case in point, the supposition that the Halflife 2 beta may have been leaked through an Outlook preview pane exploit, as other/.ers have already pointed out. Also, consider all of the hubbub about viruses shutting down public services, possibly including a transportation service and a nuclear power plant system in recent history.
Of course, this all begs the question "why the hell were the nuclear power plant, train system, and half-life build system connected to the internet in the first place?" Folks, here's a gigantic hint: software is insecure. If you want something to be secure, take it off of the fucking intarweb. The nuclear power plant just doesn't need Fark that badly. Let them read it on their PDAs.
Like the people maintaining those systems don't know better.
Sounds like Microsoft is *way* ahead of you. Observe:
1. If the user doesn't have to stop what he's doing to solve an inexplicable puzzle every few minutes, he'll be done waaaay too fast.
BSOD. Example: "What the hell? I wasn't doing anything. I was just staring at the screen. Wasn't even scrolling. Why did it crash?"
2. Obey the principle of most astonishment. Surprise the user as often as possible! Preferably with something terrifying that makes him literally fling himself out of his chair (example: the aliens in Alien Vs. Predator love to sneak up on you along walls and ceilings and suddenly let you have it from three directions -- a guaranteed excuse to press "pause" and go put on a new pair of underwear).
BSOD. Example: "Wow, these demons are really tough. (kapow, kapow; reload) I'm running out of ammo. Better run back towa-- (fling!) Where the - Fuck, it crashed!"
3. If the user screws something up, HE MUST BE PUNISHED. Usually, this means his onscreen persona (resume, spreadsheet, etc) should die a wretched, gory death, scaring the crap out of the user (see #2) and he should have to start whatever he was doing over from his last save point. This of course encourages saving documents frequently, always a good thing with Microsoft software.
BSOD. For example, there are any number of things which the user can do in control panel which will immediately illicit a Very Bad Crash (tm). This causes the user to need to restore from the last save point; however, and not to be rude, I must point out that much like anything else, this does no in fact encourage users to hit control s. Because obviously two keypresses are too much effort to safeguard your last four hours of exploratory work three hours before the paper is due.
4. If the software includes networking features, you MUST include a "taunt" feature. Allow preformatted taunts and on-the-fly taunts; both are equally fun for all. "Hey, BILL! Your powerpoint SUCKS!"
BSOD. I mean, honestly, haven't you ever gotten the impression that when one of those happens, the god damned thing is just mocking you?
5. And, finally, you have to include a few easter eggs and hidden areas. These should include a "must-have" that isn't granted to ordinary users (like, say, print preview).
This turns out to be a misfeature, though I'm not sure why. Briefly, MS offered a must-have feature in the "anti-BSOD spray" OS, better known as NT 3.5. However, MS retreated from this strategy almost a decade ago. I find your work to be insightful, but I can tell that they've been there first, and discovered it's wrong.
For non-text selections, backspace (on Windows, not Mac OS) should not delete selections. That's what the delete key is for.
What are you talking about? Can you name one major application which behaves in this fashion?
Both backspace and delete are removal keys. The difference is solely in the direction in which they remove, and when the block is pre-specified, the difference is moot.
and feel like ridiculing users will somehow socially elevate them
Speaking as a former tech support employee, I can firmly say that for me and my coworkers at the time, it was more an issue of catharsis. I've also been a telemarketer, and the flak I recieved as a technical support staffmember was far worse.
The customer acts as if it's the ISP's fault that the phone company has hosed their phone line again. The customer acts as if it's the ISP's fault that little timmy has fux0red their modem settings again. The customer acts as if it's the ISP's fault that they've forgotten their password again.
When I was tech support, I was one of two support monkies that I never saw reduced to tears in the course of about two years, and I came damned close more than once. I suspect you've no idea the sort of abuse that irate assholes subject tech support people to. So, when a user is shocked that the same thing that would happen in any application has happened here and can't be bothered to find and hit undo (I don't care what the AC said about it being a web application; undo works just fine in IE 3-6, Netscape 2-4 and Mozilla,) I think that the tech support crew is fully within their rights to be annoyed.
For the sake of professionalism, this may not be shown to the user, which the poster was careful to be clear that they did not do. Perhaps you've just got a bias towards hating technical support. Of course, if you're an Earthlink customer, I understand wholeheartedly; I'm convinced that they actually go to significant effort to find the bottom of the barrel people available.
I'd like to see a couple of said jackass tech support people be laughed at by the mechanic
Not very in touch with pop culture, are we? It is a classic truism that the service industry is mistrusted, shat on, and that the customer defends their actions by presuming that the service industry members all do this, and that they had done it first is somehow implied. You've really never heard a story about getting screwed by a car mechanic? I'd be hard pressed to find a less damning example, except maybe a used car salesman.
Hint: the services which get it the worst are the ones which people feel tied to without alternatives (utility monopolies especially,) the ones which cost the most (auto mechanics get it doubly because cars with failings often have other failings on the horizon, and the users which allow them to fail don't maintain well, leading to cascade failures which they blame on the mechanics,) and the ones they understand the least (plumbers and computer techs get this one a lot.)
It's really about insecurity. As you enter a service industry, you find that the more the customer knows about their service (on the average) the less horrible they are to you. If you can explain a delay or a problem to them, they're generally much more tolerant. The people which are the worst are the huge clueless fucks, especially the ones which don't know that they're clueless fucks, doubly especially the ones which think they're something other than a clueless fuck. Do you get shat on by tech support a lot? Could this be a hint?
(which they're unable to fix, despite the fix being a quick, five-minute change)
I'm willing to bet you're a moderate quality amateur mechanic, hence the choice of looking down this particular nose. Can you replace your sink? Can you pull up your floorboards? Have you ever installed a ducted air conditioner? Can you plan a seeding cycle? This is the sort of arrogance which nearly every professional
Slashdot Mirror
Well, maybe if the tutorials out there spent less time being condescending and more time actively presenting the real paradigms instead of flimsy confusing stuff, it wouldn't be a problem.
:D
"What's the right mouse button for?"
wrong: "it's a context-sensitive menu enabling access to control commands"
wrong: "it's like a scrapbook in which your least used situational commands are gathered and presented for your use"
right: "it's where your less common controls go. there're even rarer ones in the big menus. it works on pretty much everything. just try it out a lot; as long as you don't pick any menu items, nothing's gonna change, and you won't hurt anything."
1) Give them a simple straightforward explanation of what it does without jargon or metaphor
2) Encourage them to familiarize themselves with the control, being careful to note when such experimentation is inappropriate, even when it's never inappropriate
Not so hard. Out of curiosity, I sat through a biug chunk of the tutorial shipped with my new commodity PC; there were some things I didn't understand, and I wrote software for a living.
Perhaps hire fewer multimedia visionaries and more teachers next time you guys write intros.
You fool, 6502 assembly has no weaknesses!
Sorry, just saying luminaries wanted it banished and then not explaining left a bad taste in my mouth. I should have provided this link to a Herb Sutter article called "Can't Afford Export" (regarding getting rid of the export keyword) in the original post.
This isn't the only such article, but it's the only one I'm finding on a brief search; IIRC, I think I've also seen similar stances from Meyers and Dewhurst. Not certain, though, 'cause I can't seem to find them (yay 30 second searches.)
Comeau is the only 100% compiler; most everyone else hoses export, which is of arguable utility (and which many luminaries want to banish anyway.) For all practical purposes, Borland and Gnu have been ISO compliant for years. Intel's compiler is a monster. VC7 is actually pretty good, though VC6/eVC3 were just a mess. Sun's compiler is pretty strong. The Metaware and Gold Hills toolchains are very strong. ARM's ADS is very compliant (moreso than SDT was.) I don't honestly know the status of IBM's compilers, but I expect the best from them; same goes for SGI since their years as Cray.
Frankly, the chances that anyone has real use for anything modern compilers regularly get wrong is fairly slim. You must be *powerful* with the force to really scare G++.
That's a fairly mundane observation. I could just as easily suggest that "the fact is that if the language permits you to write anything you want, most people will write bad code."
In fact, that's quite what I'm suggesting is going on. If you have a high power toolset, you have a high powered way to write shitty code. No amount of language alteration will change that bad programmers write bad software. Just as it's not the STL's fault if you can't handle iterators, it's not the language's fault if you can't handle ranges. Don't blame your inability to write secure code on the language. Read a fucking book.
Why do you need to interview Stroustrup to learn about getting maximum from C++?
// Than are dreamt of in your philosophy."
Just do #include <algorithm>
"There are more things in heaven and earth, Horatio,
Oh, and there's more to C++ than <algorithm>
Any language that allows for someone making a living pointing out everything one shouldn't do needs more than a face-lift.
;)
It's not exactly clear to me whether I'm taking a joke on as if it were serious. That said, this is a genuine sentiment with many, so I feel my rebuttal to be germane to them if not nessecarily to you.
The problem with language is that its domain is inherently difficult. The more expressive a language is, the deeper and more powerful structures it can express. Conversely, because making a language more expressive involves adding new methods of control, the more expressive a language is, the more difficult it is. You need to know how to generate the syntax which wields the new expression mediums, etc.
The problem as I see it is that the professional field requires a significantly higher degree of skill in expression than the amateur field wants to put forth. At first this seems like a tautology, but think about it: it's fairly rare that an experienced c++ programmer suggests there needs to be a starting-over. Given that few other dominant-in-their-day languages can make that statement, I think it's worth considering that maybe C++ is as horrible as it is by necessity.
Bjarne suggests that there are more than a dozen languages named D (which is stupid; c++ *is* D, as well as P; don't try to extends jokes you don't understand, folks, kthxbye) and another half dozen named E or P. That some of them are backwards compatible and *still* haven't takjen over should be an indication that C++ is doing at least something right, despite hordes of popular, well supported contenders.
I too hate a lot of things about C++. But frankly, even though it's not even close to my first language, I can't think of another language I like more for a generic problem domain. Sure, when it's web apps it's PHP; when it's database toys or mostly UI apps it's Delphi; et cetera. Still, when it's difficult, it's c++, every single time.
Any profession that allows for a whole subprofessions making a living pointing out what one should not do has become satisfyingly deep. Things which one can just jump into and be the best at are shallow. Art, writing, engineering, mathematics, induction, politics, philosophy. Each of these has books, tomes on what you should *not* do in order to be effective. In some fields this study of counterexamples has become so important that it elevates to its own subfield of study (fallacies and converses;) arguably that's happening in CS with AntiPatterns right now.
What I think you're seeing is a language which is way ahead of the game in community support.
There's a certain subtle beauty to seeing this while it's still marked informative, not funny.
Besides, intercal > unlambda.
strncpy() has existed for quite some time, thanks. How about eliminating the coding practices that lead to buffer overflows? Oh, right, it's the *language's* fault. God forbid a language which is designed explicitly to allow you to do exactly what you want to in the nearest-to-machine way reasonable allow you to write security holes too.
Maybe instead of asking your toolset to design for you - security is a design process, not a patch process, which is why it works for qmail but not for sendmail - you should just learn to write secure code.
This paper, c++0x directions, is where he sees c++ going. :D
that Scott Meyers wrote 3 large volumes on the subject.
To bolster the point, there are a fair number of other authors which have written such series, and some of them arguably more difficult than the effective trio. Notably, Dewhurst's C++ Gotchas and Sutter's Exceptional/More Exceptional C++ are similar vein, similar-or-greater difficulty issues. None quite so well written as Meyers', though.
but this would effectively be a massive distributed DOS attack on spammers.
versus
In other words, you could host your Viagra-peddling site with a company that has a stringent no-spam policy, but a DNS lookup will point to a home user's compromised machine.
Attacking a spammer's resources only increases the spammer's impetus to steal resources. The further you push them underground, the harder they are to uproot when you get a real tool.
1) Few movies are double platter DVDs. You should be starting with 4.7g.
2) The original poster said DVD-quality. Transcoding to MPeg4 or DivX dramatically reduces size.
VCDs were thriving almost a decade ago. Compression/decompression technologies have since improved significantly, as has end-user processing power. A FoaF regularly downloads movies over broadband; takes him 15-20 min each. It takes me fifteen minutes just to get to a rental place, much less park, come back, deal with traffic, deal with rental and late fees, etc.
And while you're right that the trip to Blockbuster doesn't take ten hours, waiting for that high-school schmuck to scan two people's tapes does.
> How long would a typical DVD quality movie take to
> download?
Less time than it takes to go to Blockbuster. Less gas, too.
Very simply put, because I don't want to pay for other people's right to steal, I don't want to underwrite the industry's estimates of what their failing business model, i mean, technology is costing them, and I don't want to be told that this is the alternative to allowing a puny entertainment industry to infringe upon my rights.
It's their problem, and it doesn't come out of my pocket, amortized or not, period. This is not a socialism. Stop pretending.
http://reinvent.the.whe.el/
The TV of the future you're describing is called teh intarweb, lady. Just so you know.
no, that's just to throw you off of the trail. it's all about speaker bracelets. always has been.
just think. that night, when you were a kid... oh, but i've said too much.
Yes.
Am I the only person that thinks that messages like this are some form of steganography?
Big development effort? That can't be more than a 50-line PHP widget, given their customer records.
No, I'm not implying that it's written in PHP. I'm just suggesting that a good programmer could probably have done it in an afternoon.
ICANN shouldn't have to sue anyone over a technical aspect of the Internet. They should have the tools to simply tell Verisign to do it and have it done quickly.
/.ers have pointed out, they can just instruct the root servers to route around the damage.
They are not suing. They are, in fact, leveraging their contract - their tool - and telling verisign to get it done and have it done quickly. Specifically, 36 hours. The thing about the business world is that if they didn't make sure that they were on strong grounds, if they demanded the service be taken down and then got sued, then they'd be indemnable for whatever money verisign made up that they lost on absent sitefinder service.
ICANN is doing the right thing, in fact the very thing that we're angry that VeriSign didn't do: they're checking that their actions are correct before undertaking them. ICANN has a responsibility to be proper and careful, rather than just running around swinging its arms like a bully (which some would say that it has done in the past.)
Look, you can't please everybody: if you do it fast people will say you didn't plan, and if you plan people will say you didn't do it fast enough. Don't you think it best that they do this in the way that's most difficult for VeriSign to prevent?
It's difficult to be the good guy.
And they should also have the means to simply cut Verisign out of the loop
As has been pointed out, they have implied that they will do just that in about 36 hours if their demands aren't met. As other
(Of course, nobody seems to be pointing out that there's going to be the demand for some tremendous bandwidth and heavy servers pretty on-the-spot if they choose to do that. I find myself wondering which company will attempt to step up to the bat and steal the gold ring, if VeriSign fucks this up.)
There are actually a number of life-critical operating systems in play, a few of which have been certified by the NSA. DEC and IBM make very large zeros on this sort of thing from airports, et cetera.
Notably, lawsuits can be filed for things that just cost tremendous amounts of money. Case in point, the supposition that the Halflife 2 beta may have been leaked through an Outlook preview pane exploit, as other
Of course, this all begs the question "why the hell were the nuclear power plant, train system, and half-life build system connected to the internet in the first place?" Folks, here's a gigantic hint: software is insecure. If you want something to be secure, take it off of the fucking intarweb. The nuclear power plant just doesn't need Fark that badly. Let them read it on their PDAs.
Like the people maintaining those systems don't know better.
Sounds like Microsoft is *way* ahead of you. Observe:
1. If the user doesn't have to stop what he's doing to solve an inexplicable puzzle every few minutes, he'll be done waaaay too fast.
BSOD. Example: "What the hell? I wasn't doing anything. I was just staring at the screen. Wasn't even scrolling. Why did it crash?"
2. Obey the principle of most astonishment. Surprise the user as often as possible! Preferably with something terrifying that makes him literally fling himself out of his chair (example: the aliens in Alien Vs. Predator love to sneak up on you along walls and ceilings and suddenly let you have it from three directions -- a guaranteed excuse to press "pause" and go put on a new pair of underwear).
BSOD. Example: "Wow, these demons are really tough. (kapow, kapow; reload) I'm running out of ammo. Better run back towa-- (fling!) Where the - Fuck, it crashed!"
3. If the user screws something up, HE MUST BE PUNISHED. Usually, this means his onscreen persona (resume, spreadsheet, etc) should die a wretched, gory death, scaring the crap out of the user (see #2) and he should have to start whatever he was doing over from his last save point. This of course encourages saving documents frequently, always a good thing with Microsoft software.
BSOD. For example, there are any number of things which the user can do in control panel which will immediately illicit a Very Bad Crash (tm). This causes the user to need to restore from the last save point; however, and not to be rude, I must point out that much like anything else, this does no in fact encourage users to hit control s. Because obviously two keypresses are too much effort to safeguard your last four hours of exploratory work three hours before the paper is due.
4. If the software includes networking features, you MUST include a "taunt" feature. Allow preformatted taunts and on-the-fly taunts; both are equally fun for all. "Hey, BILL! Your powerpoint SUCKS!"
BSOD. I mean, honestly, haven't you ever gotten the impression that when one of those happens, the god damned thing is just mocking you?
5. And, finally, you have to include a few easter eggs and hidden areas. These should include a "must-have" that isn't granted to ordinary users (like, say, print preview).
This turns out to be a misfeature, though I'm not sure why. Briefly, MS offered a must-have feature in the "anti-BSOD spray" OS, better known as NT 3.5. However, MS retreated from this strategy almost a decade ago. I find your work to be insightful, but I can tell that they've been there first, and discovered it's wrong.
Now, you just have to figure out why.
For non-text selections, backspace (on Windows, not Mac OS) should not delete selections. That's what the delete key is for.
What are you talking about? Can you name one major application which behaves in this fashion?
Both backspace and delete are removal keys. The difference is solely in the direction in which they remove, and when the block is pre-specified, the difference is moot.
and feel like ridiculing users will somehow socially elevate them
Speaking as a former tech support employee, I can firmly say that for me and my coworkers at the time, it was more an issue of catharsis. I've also been a telemarketer, and the flak I recieved as a technical support staffmember was far worse.
The customer acts as if it's the ISP's fault that the phone company has hosed their phone line again. The customer acts as if it's the ISP's fault that little timmy has fux0red their modem settings again. The customer acts as if it's the ISP's fault that they've forgotten their password again.
When I was tech support, I was one of two support monkies that I never saw reduced to tears in the course of about two years, and I came damned close more than once. I suspect you've no idea the sort of abuse that irate assholes subject tech support people to. So, when a user is shocked that the same thing that would happen in any application has happened here and can't be bothered to find and hit undo (I don't care what the AC said about it being a web application; undo works just fine in IE 3-6, Netscape 2-4 and Mozilla,) I think that the tech support crew is fully within their rights to be annoyed.
For the sake of professionalism, this may not be shown to the user, which the poster was careful to be clear that they did not do. Perhaps you've just got a bias towards hating technical support. Of course, if you're an Earthlink customer, I understand wholeheartedly; I'm convinced that they actually go to significant effort to find the bottom of the barrel people available.
I'd like to see a couple of said jackass tech support people be laughed at by the mechanic
Not very in touch with pop culture, are we? It is a classic truism that the service industry is mistrusted, shat on, and that the customer defends their actions by presuming that the service industry members all do this, and that they had done it first is somehow implied. You've really never heard a story about getting screwed by a car mechanic? I'd be hard pressed to find a less damning example, except maybe a used car salesman.
Hint: the services which get it the worst are the ones which people feel tied to without alternatives (utility monopolies especially,) the ones which cost the most (auto mechanics get it doubly because cars with failings often have other failings on the horizon, and the users which allow them to fail don't maintain well, leading to cascade failures which they blame on the mechanics,) and the ones they understand the least (plumbers and computer techs get this one a lot.)
It's really about insecurity. As you enter a service industry, you find that the more the customer knows about their service (on the average) the less horrible they are to you. If you can explain a delay or a problem to them, they're generally much more tolerant. The people which are the worst are the huge clueless fucks, especially the ones which don't know that they're clueless fucks, doubly especially the ones which think they're something other than a clueless fuck. Do you get shat on by tech support a lot? Could this be a hint?
(which they're unable to fix, despite the fix being a quick, five-minute change)
I'm willing to bet you're a moderate quality amateur mechanic, hence the choice of looking down this particular nose. Can you replace your sink? Can you pull up your floorboards? Have you ever installed a ducted air conditioner? Can you plan a seeding cycle? This is the sort of arrogance which nearly every professional