When it comes to the drug cartels bringing drugs up through Mexico to the US, the main problem is mainly owing to jurisdictional issues.
The Mexico/US border is some two thousand miles long. Consequently, effectively patrolling it against smuggling is, in a word, impossible. People keep proposing fences or a wall, but those are only effective if you patrol them. Thus, there isn't a whole lot the US can do. We make a token effort, but the problem is overwhelming.
If the border were quite a bit farther south (like, across southeastern Guatemala), it would be much shorter, and the problem would be much more manageable, given US resources; but Mexico and Guatemala do not have US resources and cannot effectively patrol even this smaller border, nor do they have the resources to patrol their coasts as effectively as the US, and so they are unable to prevent the cartels from bringing in essentially unlimited quantities of contraband. Indeed, a couple of the cartels have almost as much military capacity as the Mexican government, and thus they have basically free run of most of the country (collectively; their infighting divides it into segments, so no _one_ cartel has free run of everything, which is both good and bad for Mexico, depending on how you look at it). They would not be able to manhandle the US in the same way, but they don't need to: the border is so long, they can just smuggle in whatever they want surreptitiously, once they get it into Mexico. And they can get it into Mexico relatively easily, because Mexico is not the US.
It's a difficult problem. The Mexican government would like to stop the cartels, because they cause a lot more problems for Mexico than they do for the US. (Our drug problems are annoying, but they mostly harm people who voluntarily allow themselves to be harmed, and arguably their families. The problems the cartels create in Mexico have significantly more collateral victims from the population at large. Some parts of Mexico are effectively a war zone at times, because of the cartels.) Mexico, however, doesn't have the resources to handle the problem.
In purely practical, technical terms, the US could provide the Mexican government with significant assistance in tackling the problem, which would be well worth our while to do and would be beneficial to Mexico as well, but that brings up a whole passel of interesting political complications, including a variety of internal (domestic) political forces as well as pressures both countries face from outside the hemisphere, not to mention what e.g. Guatemala would think of it. Significant care would have to be taken, or the issue could become quite thorny for everyone involved.
I've lived in the United States continuously since the mid seventies, and it's been aeons since I've heard about anyone here having a pay-by-the-megabyte internet plan. There are people still on dialup here, but to the best of my knowledge everybody (who has internet at all) has theoretically unlimited internet (though it can be a bit congested during peak hours and not provide the advertised speed).
My main objection to HTML5 is that it is a step back toward the bad old days of HTML4 when parsing the markup was a royal pain in the hind end. XHTML's concept of well-formedness is so immensely useful, I cannot imagine anyone who understands the implications ever wanting to go back to the horrible morass of SGML-based markup. XML-based markup is so much easier to manage, both for the content creator (e.g., web developers) and also for software developers (browsers, editors, server-side stuff, indexers, anything that works with markup).
I have no intrinsic objection to killing off Flash, although I also wouldn't mind having a nickel for every technology that has been promoted with the claim that it would do so. I could sort of understand how someone *might* think Silverlight would kill off Flash, if said someone did not understand about backward compatibility. (Hint: why didn't Itanium kill off 32-bit processors?) I do not, however, understand how anyone could suppose that HTML5 could kill off flash.
> If you had actually tried to install Windows from a CD, you would be terrified too.
I'm a network administrator. I've installed Windows more time than most people have checked their email.
> In all probability, you would be a Ubuntu user.
Debian stable, currently. I used Ubuntu briefly when it was new (Warty and Hoary; because Debian stable at that time was *extremely* out of date), but when Sarge came out I switched back over to Debian and haven't had a reason to move away from it since. Other distributions I've used in the past include RedHat, Storm Linux, Mandrake, Gentoo, Knoppix, CentOS, an obscure application-specific Slackware derivative, FreeBSD, and BeOS, plus DOS 3, 5, and 6 and every major Windows version except NT3 and Eight (waiting for SP1 on that), plus Mac OS X. Oh, I messed around with QNX for a couple of days once also. That's just ones I've installed, and I'm probably forgetting some. (OSes I've worked with but not installed include Mac System 6 and 7, MacOS 8 and 9, and VMS 6 and 7. I would NOT attempt to install VMS without assistance, unless I could find clear and detailed instructions for it on the internet -- which is possible; I haven't looked.)
> Support for hardware on the Windows install disk is so grim, it will probably not be useable.
Actually, that has gotten a lot better. Modern Windows systems can often detect enough about the video hardware to use the best resolution and color depth your hardware supports. It used to be you could count on being stuck in 16-color mode for the first while. Also, network hardware is MUCH more likely to work out of the box with Windows now than it was a few versions ago, which is REALLY nice, because it means you can actually download any additional drivers you need from the manufacturer's website, instead of having to use another system to track them down and then figure out how to transfer them over when they typically were too large to fit on a floppy. On the rare occasion that the network card doesn't work, USB (another thing that never used to work out of the box) does, so I can transfer the drivers that way. I used to just about take for granted that I'd probably have to yank the hard drive after a Windows reinstall and plug it into a different computer just to save some drivers on it, then put it back in the computer it belonged in, boot, and finally actually install the drivers. I haven't had to do that lately.
It's still often necessary to download *some* driver or another from the manufacturer's website. Printers are a common offender.
> I hate upgrading. Gotta [do a whole bunch of fooling around].
That's because you use Windows, which has two shortcomings in this area: First, it doesn't have proper package management (which means the OS can't automatically check for new versions of your applications that will work better with the new OS and handle upgrading those too at the same time), and second the Windows registry is an insane mess that was very clearly NOT designed with upgrades in mind.
When I upgrade my Debian system, I don't have those problems. Typically I keep Firefox open and continue reading Slashdot or Wikipedia or something while the upgrade happens in the background. As long as I don't *close* the browser window, the upgrade-in-progress doesn't usually interfere with using it. (If I closed the browser, I'd likely have to wait for the upgrade to finish before I could launch it again, since the various libraries that it's linked against aren't necessarily all upgraded at exactly the same moment as the browser itself.)
When the upgrade is done, I do have to reboot once, assuming the upgrade included a kernel update, which it generally does. It would be kind of nice if that reboot weren't necessary. It's annoying to have to open all my windows again. I also have to reboot after power outages, which is similarly annoying. Somebody should do something about that.
The only applications I'd have to reinstall would be ones I compiled from source, and even then only if they depend on libraries that were updated as part of the upgrade. Those have to be recompiled.
It's not 1996 any more. Nobody in the developed world still pays by the kilobyte, let alone by the bit. (Unless you count SMS -- a few cellphone users still have caps on that and get charged per-message if they go over -- but that's not applicable here, as we were talking about web pages.)
Also, a couple of extra CSS properties only adds up a few dozen bits, which in the scheme of things is not much. I can easily more than make up for it by NOT loading down every web page with 817 megabytes of entirely gratuitous Javascript, Flash, and whatnot.
> > it'll require doing things that annoy large volume customers. > Since when has MS been averse to doing things > that annoy large volumes of paying customers
Do you not see a difference between "large volume customers" (which, admittedly, should really be hyphenated) versus "large volumes of... customers"?
For Microsoft, there's a very big difference. Microsoft certainly doesn't mind annoying large volumes of their customers, as long as they're NOT the large-volume customers. This distinction explains, among other things, why Automatic Updates cannot be set to go ahead and install but wait up to 24 hours for the user to shut the computer down. Normal people are annoyed, because they don't want whatever they're doing with the computer to be interrupted. Microsoft's official answer is to set the updates to happen in the middle of the night. This answer satisfies Microsoft's large-volume customers, because they all leave all their computers running all night for no reason while the building is locked. Normal people shut the computer down at night, so then they get interrupted for a mandatory restart during their working day, but normal people don't matter, because they're not large-volume customers. This could be solved by adding a "wait up to 24 hours before forcing a restart" setting, but Microsoft can't be bothered to do that because the large-volume customers don't care.
Summary says: > 'We don't know what level of marijuana impairs a driver.'
How about "any amount that causes the loser to drive in such a manner as to get pulled over and then follow it up by failing a sobriety test"?
Too obvious?
While we're at it, if the moron is so addicted to SMS that he tries to text while driving and gets pulled over for reckless driving, and then he continues trying to text during the sobriety test and fails it as a result, that should count as impairment too. Take away his driver's license, impound the vehicle, and require him to attend six months of counseling for his addiction and test clean if he ever wants to get a driver's license again.
> 3. Nethack is to 'real' gamers what Excel is to real programmers:)
Actually, I think it's more like machine code. In fact, it bears certain striking resemblances to writing machine code in hex for a drum memory computer, a la The Story of Mel.
For example, you know the part of the story where the narrator discovers that Mel's code actually deliberately abuses counter overflow? Experienced NetHack players actually do exactly that, in order to achieve various Stupid Ascension Tricks.
When I read the headline, "Click-to-Play Plugin Blocks", I was thinking that plugin content would be blocked from doing anything unless the user clicks a play button. Just like FlashBlock, in other words. That would actually be a good thing. A good change, in a new version of Firefox: I might've fainted.
But no, what it actually means is this: > Mozilla will now prompt Firefox users on Windows with old versions of Adobe Reader...
Oh, yes, please.
We need this because Adobe Reader doesn't already prompt every single user who has it installed to the effect that they need to upgrade it, a bare minimum of three per hour. We definitely need our web browser to bug us about this also, otherwise we might not know that three new versions of Adobe Reader were released during the time it took us to download and install the version we currently have. Well, I mean, okay, in theory we'd _know_, but without this extra reminder we might occasionally go up to fifteen minutes at a time without _thinking_ about it. Mozilla must protect us from that horrific fate.
> How many people actually 'upgrade' without buying a new computer?
Well, let's see, there's Debian users, Gentoo users,...
Oh, wait, you meant *Windows* users? Upgrade the OS?
Haha. That would mean they'd have to *install* the new version of the operating system. Like, from a CD. Do you have any idea how terrified Windows users are of installing anything (that doesn't come from an obviously scammy website or email message)?
The people who aren't afraid to try to install an operating system all either switched to Linux years ago or else make their living doing Windows support of one kind or another.
> if you dont mind blocky distortion around your line art too You're thinking of JPEG.
PNG is lossless (for purely two-dimensional images).
GIF is theoretically lossless (in the sense that saving repeatedly does not degrade the image further), but it (generally, in practice) means reducing your colorspace to 8-bit, which entails a significant amount of information loss for most images. Though it would be fine for something with almost no color anyway, e.g., xkcd.
Of course, if you want to preserve editing features (e.g., layers and masks), then you're going to save in a format that supports such things, probably either.psd or.xcf (or a compressed version thereof) in most cases.
But PNG is fine for distribution to people who are just going to look at the image and not edit it -- e.g., for display on a website, or for printing.
Indeed. If I were a browser developer, I'd be tempted to treat all the major browser-specific prefixes, including my own, identically: render as if the prefix weren't specified at all.
Web developers _should_, in theory, use the prefix for testing before the thing is standardized and then switch over to unprefixed versions subsequently. In practice, that doesn't always happen, partly because after it's standardized not all browsers support the unprefixed version right away but mainly because changing all your code is extra work, and furthermore the web developer may not currently be updating the code in question when the feature finally becomes standard.
What I usually do in practice, as a web content developer, is list both the expected but not yet supported un-prefixed property and the prefixed one that is already supported (sometimes more than one, if multiple browser engines have each their own prefixed but otherwise identical version of it, which sometimes happens), with the same value. Then when browsers eventually implement the standard unprefixed property, Everything Just Works -- including for people who are using old browsers that only support the prefixed property. (Obviously, even older browsers that don't support it at all get a rendering without that feature. There's only so much you can do.) This approach only works if you adopt new properties sufficiently late in their cycle that you can reasonably predict what the property is probably going to look like when it's finally standardized. For me, that's usually the case.
Selectors are somewhat more problematic. In many cases have to wait until all the browsers you really care about support them (although when implementing purely aesthetic cosmetic features you can just let non-supporting browsers hang in the breeze).
> When machines start translating languages on the fly, people > will stop learning other languages and that's a bad thing.
Relax. When they figure out how to make a machine that can translate from Japanese to English (or vice versa) and have the results be even remotely intelligible (nevermind about grammatically correct or accurate), we won't *need* humans any more, because the machines will be able to do everything for themselves, including fine art and great literature.
Well, see, it's like this. As a web developer, I of course have a burning fiery passionate hatred of IE in general.
Importantly, however, I hate the older versions even more than the newer ones. IE8 is bad, yes, but IE7 is a million times worse, and IE6 is basically a cross between cancer and the holocaust. If you so much as mention any version before that, I will personally hunt you down and force you to watch ten thousand continuous hours of back-to-back car dealership commercials.
Thus, I am constantly on the edge of my seat wanting new versions of IE to come out, even though I know I am going to hate them. It's not because I want people to use IE but rather because I want people to upgrade off of their old versions sooner than is humanly possible. When should IE8 users upgrade to IE10? This morning would be good. Last month would be even better. The ideal time would be some time before the IE8 beta was actually released. 1997 would be fantastic. IE7 users should upgrade even sooner. Any time before I was born would be good, that way I'd never have to know about IE7 in my lifetime. If IE6 users could upgrade during the Achaemenid dynasty, that would be great. The sooner the better.
So it actually bothers me, for example, that IE10 doesn't support Windows XP. I don't use IE, and I don't use Windows XP, but I want the new version of IE to run on Windows XP. Frankly, I want it to be distributed to Windows XP users via Automatic Update. If it were possible to get IE10 into XP SP2, I'd be in favor of that.
That's what this is for: <meta http-equiv="X-UA-Compatible" content="IE=Edge"/>
Yes, granted, it SHOULD be the default, because there are only like three websites in the whole world that actually render _better_ on older versions of IE than on the newer version, and that's only because they haven't been updated in a full decade. We shouldn't HAVE to declare in our website markup that our site is more up-to-date than IE, because it's rare to find a site that isn't (excluding corporate intranets, but those are only accessed from systems that the site admins control anyway, so they still have IE4 or whatever antedeluvian version they were originally written around back in the punched-card era).
Nonetheless, putting that one little thing in your html head makes the entire broken document icon problem go away. The only downside, if it even counts as a downside, is that you are then committed to keeping your site at least as up-to-date as Internet Explorer. Yeah, that's gonna be tough. I might have to update my site, like, every three or four YEARS (unless I actually code to standards in the first place -- imagine that).
> Maybe the Slashdot developers will even > have time to implement UTF-8 support
What would be the point? The lameness filter is going to strip out everything but printable ASCII anyway. (In general, this is mainly a good thing, because it filters out a great deal of, in a word, lameness. Admittedly, it occasionally annoys me that I can't include one short foreign word in an otherwise English-language post.)
The correct approach is to stop trying to flatten the grading scale into only two categories (pass and fail). That's stupid. Give up on that and just report each student's percentage. If Hanako gets an 84 and John gets a 73 and Shu'Dazzi gets a 61, report their scores as 84, 73, and 61. Sorted. Whether that's a good enough score is really up to the college or prospective employer or whoever.
I don't particularly like letter grades either, for much the same reason. Five categories is better than two, but it's still pretty undiscriminating. You end up with students who don't bother to study for the last couple of tests because they realize it can't change their grade. Just report everything as a percentage, and let the students do as well as they can (or as well as they are willing to bother with, which is how you get a score like 61).
There are several reasonably good ways to accomplish this, but the most obvious is to give them the end result image in a format that does not support layers (e.g., png). The students, of course, must turn in their assignment in layered format, on the "show your work" principle.
This doesn't prevent students from working together, of course.
You're confusing intelligence with educatedness. They are distinct concepts. Young children are much more intelligent (on average) than adults, but much less educated. Today we are (on average) much better educated than we were a thousand years ago, but we are significantly less intelligent.
If you think about it in terms of genetics, it's actually rather obvious that given enough time this will happen in all populations (well, all populations that neither go extinct nor have significant genetic material introduced via immigration). Physical health also degrades in a population over time. It's called inbreeding -- or, more generally, entropy.
> The Enigma was used by the Germans, not the British.
Yes, but they didn't just use it _in Germany_, and it is reasonable to imagine that they may have had people in England. I was not aware that we knew which side sent this particular message.
Slashdot Mirror
> It has continued for fifty years so far.
Fifty years, are you kidding? More like four thousand years.
When it comes to the drug cartels bringing drugs up through Mexico to the US, the main problem is mainly owing to jurisdictional issues.
The Mexico/US border is some two thousand miles long. Consequently, effectively patrolling it against smuggling is, in a word, impossible. People keep proposing fences or a wall, but those are only effective if you patrol them. Thus, there isn't a whole lot the US can do. We make a token effort, but the problem is overwhelming.
If the border were quite a bit farther south (like, across southeastern Guatemala), it would be much shorter, and the problem would be much more manageable, given US resources; but Mexico and Guatemala do not have US resources and cannot effectively patrol even this smaller border, nor do they have the resources to patrol their coasts as effectively as the US, and so they are unable to prevent the cartels from bringing in essentially unlimited quantities of contraband. Indeed, a couple of the cartels have almost as much military capacity as the Mexican government, and thus they have basically free run of most of the country (collectively; their infighting divides it into segments, so no _one_ cartel has free run of everything, which is both good and bad for Mexico, depending on how you look at it). They would not be able to manhandle the US in the same way, but they don't need to: the border is so long, they can just smuggle in whatever they want surreptitiously, once they get it into Mexico. And they can get it into Mexico relatively easily, because Mexico is not the US.
It's a difficult problem. The Mexican government would like to stop the cartels, because they cause a lot more problems for Mexico than they do for the US. (Our drug problems are annoying, but they mostly harm people who voluntarily allow themselves to be harmed, and arguably their families. The problems the cartels create in Mexico have significantly more collateral victims from the population at large. Some parts of Mexico are effectively a war zone at times, because of the cartels.) Mexico, however, doesn't have the resources to handle the problem.
In purely practical, technical terms, the US could provide the Mexican government with significant assistance in tackling the problem, which would be well worth our while to do and would be beneficial to Mexico as well, but that brings up a whole passel of interesting political complications, including a variety of internal (domestic) political forces as well as pressures both countries face from outside the hemisphere, not to mention what e.g. Guatemala would think of it. Significant care would have to be taken, or the issue could become quite thorny for everyone involved.
I've lived in the United States continuously since the mid seventies, and it's been aeons since I've heard about anyone here having a pay-by-the-megabyte internet plan. There are people still on dialup here, but to the best of my knowledge everybody (who has internet at all) has theoretically unlimited internet (though it can be a bit congested during peak hours and not provide the advertised speed).
My main objection to HTML5 is that it is a step back toward the bad old days of HTML4 when parsing the markup was a royal pain in the hind end. XHTML's concept of well-formedness is so immensely useful, I cannot imagine anyone who understands the implications ever wanting to go back to the horrible morass of SGML-based markup. XML-based markup is so much easier to manage, both for the content creator (e.g., web developers) and also for software developers (browsers, editors, server-side stuff, indexers, anything that works with markup).
I have no intrinsic objection to killing off Flash, although I also wouldn't mind having a nickel for every technology that has been promoted with the claim that it would do so. I could sort of understand how someone *might* think Silverlight would kill off Flash, if said someone did not understand about backward compatibility. (Hint: why didn't Itanium kill off 32-bit processors?) I do not, however, understand how anyone could suppose that HTML5 could kill off flash.
> If you had actually tried to install Windows from a CD, you would be terrified too.
I'm a network administrator. I've installed Windows more time than most people have checked their email.
> In all probability, you would be a Ubuntu user.
Debian stable, currently. I used Ubuntu briefly when it was new (Warty and Hoary; because Debian stable at that time was *extremely* out of date), but when Sarge came out I switched back over to Debian and haven't had a reason to move away from it since. Other distributions I've used in the past include RedHat, Storm Linux, Mandrake, Gentoo, Knoppix, CentOS, an obscure application-specific Slackware derivative, FreeBSD, and BeOS, plus DOS 3, 5, and 6 and every major Windows version except NT3 and Eight (waiting for SP1 on that), plus Mac OS X. Oh, I messed around with QNX for a couple of days once also. That's just ones I've installed, and I'm probably forgetting some. (OSes I've worked with but not installed include Mac System 6 and 7, MacOS 8 and 9, and VMS 6 and 7. I would NOT attempt to install VMS without assistance, unless I could find clear and detailed instructions for it on the internet -- which is possible; I haven't looked.)
> Support for hardware on the Windows install disk is so grim, it will probably not be useable.
Actually, that has gotten a lot better. Modern Windows systems can often detect enough about the video hardware to use the best resolution and color depth your hardware supports. It used to be you could count on being stuck in 16-color mode for the first while. Also, network hardware is MUCH more likely to work out of the box with Windows now than it was a few versions ago, which is REALLY nice, because it means you can actually download any additional drivers you need from the manufacturer's website, instead of having to use another system to track them down and then figure out how to transfer them over when they typically were too large to fit on a floppy. On the rare occasion that the network card doesn't work, USB (another thing that never used to work out of the box) does, so I can transfer the drivers that way. I used to just about take for granted that I'd probably have to yank the hard drive after a Windows reinstall and plug it into a different computer just to save some drivers on it, then put it back in the computer it belonged in, boot, and finally actually install the drivers. I haven't had to do that lately.
It's still often necessary to download *some* driver or another from the manufacturer's website. Printers are a common offender.
> I hate upgrading. Gotta [do a whole bunch of fooling around].
That's because you use Windows, which has two shortcomings in this area: First, it doesn't have proper package management (which means the OS can't automatically check for new versions of your applications that will work better with the new OS and handle upgrading those too at the same time), and second the Windows registry is an insane mess that was very clearly NOT designed with upgrades in mind.
When I upgrade my Debian system, I don't have those problems. Typically I keep Firefox open and continue reading Slashdot or Wikipedia or something while the upgrade happens in the background. As long as I don't *close* the browser window, the upgrade-in-progress doesn't usually interfere with using it. (If I closed the browser, I'd likely have to wait for the upgrade to finish before I could launch it again, since the various libraries that it's linked against aren't necessarily all upgraded at exactly the same moment as the browser itself.)
When the upgrade is done, I do have to reboot once, assuming the upgrade included a kernel update, which it generally does. It would be kind of nice if that reboot weren't necessary. It's annoying to have to open all my windows again. I also have to reboot after power outages, which is similarly annoying. Somebody should do something about that.
The only applications I'd have to reinstall would be ones I compiled from source, and even then only if they depend on libraries that were updated as part of the upgrade. Those have to be recompiled.
> Forgot ME?
Different entirely. 95 / 95 OSR2 / 98 / 98SE / Me, and then that product line was discontinued.
> Which means users have to pay by the bit
It's not 1996 any more. Nobody in the developed world still pays by the kilobyte, let alone by the bit. (Unless you count SMS -- a few cellphone users still have caps on that and get charged per-message if they go over -- but that's not applicable here, as we were talking about web pages.)
Also, a couple of extra CSS properties only adds up a few dozen bits, which in the scheme of things is not much. I can easily more than make up for it by NOT loading down every web page with 817 megabytes of entirely gratuitous Javascript, Flash, and whatnot.
> > it'll require doing things that annoy large volume customers.
... customers"?
> Since when has MS been averse to doing things
> that annoy large volumes of paying customers
Do you not see a difference between "large volume customers" (which, admittedly, should really be hyphenated) versus "large volumes of
For Microsoft, there's a very big difference. Microsoft certainly doesn't mind annoying large volumes of their customers, as long as they're NOT the large-volume customers. This distinction explains, among other things, why Automatic Updates cannot be set to go ahead and install but wait up to 24 hours for the user to shut the computer down. Normal people are annoyed, because they don't want whatever they're doing with the computer to be interrupted. Microsoft's official answer is to set the updates to happen in the middle of the night. This answer satisfies Microsoft's large-volume customers, because they all leave all their computers running all night for no reason while the building is locked. Normal people shut the computer down at night, so then they get interrupted for a mandatory restart during their working day, but normal people don't matter, because they're not large-volume customers. This could be solved by adding a "wait up to 24 hours before forcing a restart" setting, but Microsoft can't be bothered to do that because the large-volume customers don't care.
Summary says:
> 'We don't know what level of marijuana impairs a driver.'
How about "any amount that causes the loser to drive in such a manner as to get pulled over and then follow it up by failing a sobriety test"?
Too obvious?
While we're at it, if the moron is so addicted to SMS that he tries to text while driving and gets pulled over for reckless driving, and then he continues trying to text during the sobriety test and fails it as a result, that should count as impairment too. Take away his driver's license, impound the vehicle, and require him to attend six months of counseling for his addiction and test clean if he ever wants to get a driver's license again.
> 3. Nethack is to 'real' gamers what Excel is to real programmers :)
Actually, I think it's more like machine code. In fact, it bears certain striking resemblances to writing machine code in hex for a drum memory computer, a la The Story of Mel.
For example, you know the part of the story where the narrator discovers that Mel's code actually deliberately abuses counter overflow? Experienced NetHack players actually do exactly that, in order to achieve various Stupid Ascension Tricks.
When I read the headline, "Click-to-Play Plugin Blocks", I was thinking that plugin content would be blocked from doing anything unless the user clicks a play button. Just like FlashBlock, in other words. That would actually be a good thing. A good change, in a new version of Firefox: I might've fainted.
But no, what it actually means is this:
> Mozilla will now prompt Firefox users on Windows with old versions of Adobe Reader...
Oh, yes, please.
We need this because Adobe Reader doesn't already prompt every single user who has it installed to the effect that they need to upgrade it, a bare minimum of three per hour. We definitely need our web browser to bug us about this also, otherwise we might not know that three new versions of Adobe Reader were released during the time it took us to download and install the version we currently have. Well, I mean, okay, in theory we'd _know_, but without this extra reminder we might occasionally go up to fifteen minutes at a time without _thinking_ about it. Mozilla must protect us from that horrific fate.
> How many people actually 'upgrade' without buying a new computer?
...
Well, let's see, there's Debian users, Gentoo users,
Oh, wait, you meant *Windows* users? Upgrade the OS?
Haha. That would mean they'd have to *install* the new version of the operating system. Like, from a CD. Do you have any idea how terrified Windows users are of installing anything (that doesn't come from an obviously scammy website or email message)?
The people who aren't afraid to try to install an operating system all either switched to Linux years ago or else make their living doing Windows support of one kind or another.
By this logic, Windows 2000 would be good, and NT4 would be bad. Is that really an assertion you're prepared to make?
> if you dont mind blocky distortion around your line art too
.psd or .xcf (or a compressed version thereof) in most cases.
You're thinking of JPEG.
PNG is lossless (for purely two-dimensional images).
GIF is theoretically lossless (in the sense that saving repeatedly does not degrade the image further), but it (generally, in practice) means reducing your colorspace to 8-bit, which entails a significant amount of information loss for most images. Though it would be fine for something with almost no color anyway, e.g., xkcd.
Of course, if you want to preserve editing features (e.g., layers and masks), then you're going to save in a format that supports such things, probably either
But PNG is fine for distribution to people who are just going to look at the image and not edit it -- e.g., for display on a website, or for printing.
Indeed. If I were a browser developer, I'd be tempted to treat all the major browser-specific prefixes, including my own, identically: render as if the prefix weren't specified at all.
Web developers _should_, in theory, use the prefix for testing before the thing is standardized and then switch over to unprefixed versions subsequently. In practice, that doesn't always happen, partly because after it's standardized not all browsers support the unprefixed version right away but mainly because changing all your code is extra work, and furthermore the web developer may not currently be updating the code in question when the feature finally becomes standard.
What I usually do in practice, as a web content developer, is list both the expected but not yet supported un-prefixed property and the prefixed one that is already supported (sometimes more than one, if multiple browser engines have each their own prefixed but otherwise identical version of it, which sometimes happens), with the same value. Then when browsers eventually implement the standard unprefixed property, Everything Just Works -- including for people who are using old browsers that only support the prefixed property. (Obviously, even older browsers that don't support it at all get a rendering without that feature. There's only so much you can do.) This approach only works if you adopt new properties sufficiently late in their cycle that you can reasonably predict what the property is probably going to look like when it's finally standardized. For me, that's usually the case.
So, for example:
div.someclass {
-moz-awesome-new-feature: 3px 3px 3px #294D4A #FFE6BC;
-webkit-awesome-new-feature: 3px 3px 3px #294D4A #FFE6BC;
awesome-new-feature: 3px 3px 3px #294D4A #FFE6BC;
}
Selectors are somewhat more problematic. In many cases have to wait until all the browsers you really care about support them (although when implementing purely aesthetic cosmetic features you can just let non-supporting browsers hang in the breeze).
> When machines start translating languages on the fly, people
> will stop learning other languages and that's a bad thing.
Relax. When they figure out how to make a machine that can translate from Japanese to English (or vice versa) and have the results be even remotely intelligible (nevermind about grammatically correct or accurate), we won't *need* humans any more, because the machines will be able to do everything for themselves, including fine art and great literature.
> What's so complex about "burning hatred"?
Well, see, it's like this. As a web developer, I of course have a burning fiery passionate hatred of IE in general.
Importantly, however, I hate the older versions even more than the newer ones. IE8 is bad, yes, but IE7 is a million times worse, and IE6 is basically a cross between cancer and the holocaust. If you so much as mention any version before that, I will personally hunt you down and force you to watch ten thousand continuous hours of back-to-back car dealership commercials.
Thus, I am constantly on the edge of my seat wanting new versions of IE to come out, even though I know I am going to hate them. It's not because I want people to use IE but rather because I want people to upgrade off of their old versions sooner than is humanly possible. When should IE8 users upgrade to IE10? This morning would be good. Last month would be even better. The ideal time would be some time before the IE8 beta was actually released. 1997 would be fantastic. IE7 users should upgrade even sooner. Any time before I was born would be good, that way I'd never have to know about IE7 in my lifetime. If IE6 users could upgrade during the Achaemenid dynasty, that would be great. The sooner the better.
So it actually bothers me, for example, that IE10 doesn't support Windows XP. I don't use IE, and I don't use Windows XP, but I want the new version of IE to run on Windows XP. Frankly, I want it to be distributed to Windows XP users via Automatic Update. If it were possible to get IE10 into XP SP2, I'd be in favor of that.
That's what this is for: />
<meta http-equiv="X-UA-Compatible" content="IE=Edge"
Yes, granted, it SHOULD be the default, because there are only like three websites in the whole world that actually render _better_ on older versions of IE than on the newer version, and that's only because they haven't been updated in a full decade. We shouldn't HAVE to declare in our website markup that our site is more up-to-date than IE, because it's rare to find a site that isn't (excluding corporate intranets, but those are only accessed from systems that the site admins control anyway, so they still have IE4 or whatever antedeluvian version they were originally written around back in the punched-card era).
Nonetheless, putting that one little thing in your html head makes the entire broken document icon problem go away. The only downside, if it even counts as a downside, is that you are then committed to keeping your site at least as up-to-date as Internet Explorer. Yeah, that's gonna be tough. I might have to update my site, like, every three or four YEARS (unless I actually code to standards in the first place -- imagine that).
> Maybe the Slashdot developers will even
> have time to implement UTF-8 support
What would be the point? The lameness filter is going to strip out everything but printable ASCII anyway. (In general, this is mainly a good thing, because it filters out a great deal of, in a word, lameness. Admittedly, it occasionally annoys me that I can't include one short foreign word in an otherwise English-language post.)
The correct approach is to stop trying to flatten the grading scale into only two categories (pass and fail). That's stupid. Give up on that and just report each student's percentage. If Hanako gets an 84 and John gets a 73 and Shu'Dazzi gets a 61, report their scores as 84, 73, and 61. Sorted. Whether that's a good enough score is really up to the college or prospective employer or whoever.
I don't particularly like letter grades either, for much the same reason. Five categories is better than two, but it's still pretty undiscriminating. You end up with students who don't bother to study for the last couple of tests because they realize it can't change their grade. Just report everything as a percentage, and let the students do as well as they can (or as well as they are willing to bother with, which is how you get a score like 61).
There are several reasonably good ways to accomplish this, but the most obvious is to give them the end result image in a format that does not support layers (e.g., png). The students, of course, must turn in their assignment in layered format, on the "show your work" principle.
This doesn't prevent students from working together, of course.
You're confusing intelligence with educatedness. They are distinct concepts. Young children are much more intelligent (on average) than adults, but much less educated. Today we are (on average) much better educated than we were a thousand years ago, but we are significantly less intelligent.
If you think about it in terms of genetics, it's actually rather obvious that given enough time this will happen in all populations (well, all populations that neither go extinct nor have significant genetic material introduced via immigration). Physical health also degrades in a population over time. It's called inbreeding -- or, more generally, entropy.
> The Enigma was used by the Germans, not the British.
Yes, but they didn't just use it _in Germany_, and it is reasonable to imagine that they may have had people in England. I was not aware that we knew which side sent this particular message.
It's too late for that. At this point, trying to get Gnome 1.x to compile against modern system libraries would be an exercise in frustration.