The significant fact to my thesis was that they will delete your information.
No, they won't. They will hide it, and can bring it back on demand.
If you don't get the difference, then you need to talk to a few more of those "most pedantic privacy advocates" and find out how many all too serious problems happen to ordinary people every year because "deleted" information such as personal details or account numbers wasn't actually deleted at all. Perhaps you've heard of identity theft? Stalking? Credit card fraud?
Your friends can post pictures of you with your name in any media ranging from Flickr to a bathroom stall.
Actually, systematically collecting a database of personally identifiable information about anyone without their consent is illegal in some places, and comes with significant limitations on what you're allowed to do in many more. The fact that it's legal at all is a problem, as is the fact that not everywhere gives people a right to force people storing data about them to delete it — properly — if there isn't a good reason to keep it around.
Society currently fails to recognise the dangers inherent in allowing this sort of behaviour. It will, before long, when enough people get hurt. It's just a shame that until everyone understands the idea that just because we can do something it doesn't mean we should, a lot of people are going to get hurt needlessly.
In what way was my post incorrect? Accessing any of the content hosted on Facebook so you can see how it works requires a user account, or at least it did at the time I signed up. There is no way one could, to give the obvious example, conduct a simple experiment to see how much information they really collect without signing up first (or working with someone else who did so).
I'm sorry, but every single thing you wrote in your first paragraph is objectively, factually incorrect — other than the fact that you're not so sure about some of it, but that just makes you wrong.
You didn't do your research, and want to complain about it after the fact while taking no personal responsibility.
Ah, another genius who not only has time to read all the detailed agreements he ever sees and understand them, but is also blessed with prescience that allows him to interpret the intentions of other parties ahead of time.
So how do you propose I should do some research on the nature of the web site, given that to gain any access at all requires creating an account, and that it is the concessions made when creating the account that is in question here?
You can only take responsibility for something you have some reasonable control over. There was no realistic way to know what would happen in this case.
I think the view you present is perhaps too simple. There is no one check box that says a patent is "legal" or not. There is a question of whether a patent has been granted by a patent awarding body. There is a question of how widely such a patent is recognised. There is a question of whether infringing claims in an awarded patent is legally actionable in any given jurisdiction.
It is perfectly rational to argue that a patent was awarded lawfully by a recognised body, regardless of whether any legal action could successfully be brought against someone infringing it in a given jurisdiction at the time. You could also make a rational case that if a patent was properly awarded by the relevant body, no previous test had found it unenforceable in court, and infringement is actionable, then there has been no change in the legal situation.
Now, I'm not offering any view on the morality of taking such a position, nor claiming that the law would necessarily support it, nor necessarily disagreeing with the general ethical principle that retroactive laws are a bad idea. I'm just saying that a good lawyer could surely build a reasonable case, and it's nowhere near as safe or obvious to ignore dubious software patents as you seem to be suggesting.
I find the Facebook privacy stories frustrating because they seem to always ignore one thing - in almost every case the Facebook user decides how much information to make public, to whom, and which applications to install.
Facebook are quite happy to collect information on anyone who has ever been a user, including identifying them in photographs, even if they closed their account immediately after discovering that the site is one big invasion of privacy. Facebook offer no mechanism for ex-users to permanently delete such information, nor to prevent others continuing to provide it after a user cancels their account (despite the fact that this is almost certainly illegal in many jurisdictions).
So what next? Anyone whose friends group has ever mentioned a product on someone's wall is consenting to to any image of them tagged in a photograph by someone else without their knowledge being used in advertising? Anyone who once mentioned something privately to their friends in order to criticise it gets their face used to promote that thing to the world?
There is just no excuse for this. It's exactly why I am the guy who quit Facebook almost as soon as I'd joined it. Facebook, like Google, is one of the biggest dangers to modern society. Society just hasn't realised it yet, and lets them get away with stuff because they present the appearance of a useful service. Pandora's box ought to be required reading in schools.
Why is CMYK so important, don't the new, modern presses use RGB natively now?
It's not that simple. Sure, many places can convert RGB material if that's what you supply, but it doesn't have the same reliability and flexibility as CMYK for professional quality work. This is pretty obvious if you think about it, given both that CMYK has an extra "axis" and that CMYK is what will typically be used for the physical printing itself.
On a different but still relevant note, it's also considerably cheaper to use just one or two spot colours rather than a full colour print if you don't need the whole range. However, if you're using specific choices from something like a Pantone swatch, you really need some software support to use those effectively. As I understand it, the GIMP doesn't do this yet, either.
For the record, I don't work in media, but I have done publicity work for large organisations involving dealing with external print shops. I guess you might call my perspective a "semi-professional" point of view.
Your point about software patents outside the US is well taken, and indeed even in the US I would think there was a huge amount of prior art to invalidate all but the most specific claims.
However, I'm afraid things aren't as simple elsewhere as your retroactive law argument suggests. This is why certain companies are trying to get what are effectively software patents onto the books in more permissive jurisdictions within Europe (I seem to recall Ireland being mentioned as a favourite target for this, though I'm not sure what makes it more appealing). The law doesn't currently say patents can't be awarded in category X, it just says which ones can be. However, if you can get something onto the books that wouldn't actually be enforceable under the patent laws in European countries today, but big business succeeds in its repeated attempts to legitimise software patents somewhere down the line, then it won't be a retroactive measure because the patent was approved earlier. As with many things involving big business and the law, it's not necessarily the spirit that counts, but how you play the game and what technicalities you know about the rules.:-(
While I can certainly understand why you might want to work that way yourself, I think there might be a few practical difficulties with applying your reasoning more generally.
For one thing, you mentioned effectively training the user on demand when they need to do new things, but both before-the-fact training and after-the-fact on-line help or support calls work much better if the context is consistent. I notice that the Microsoft Office team — who, for all their sins, are pretty careful about their user interface and actively collect feedback — have moved in exactly the opposite direction. The magic hiding menus feature was pretty much universally slammed, and with Office 2007 they've gone for something much less customisable than before. And despite the bitching on Slashdot because they're Microsoft and they changed something, I've yet to meet someone in the real world who didn't say they preferred the new version after a little time getting used to it.
A related point is that for all the UI customisability in many modern applications, almost no-one actually uses it. Simple things like setting up styles and templates in word processors or presentation packages are ignored in favour of ad-hoc formatting. Personally, I think this has a lot to do with the fact that UIs make it easier to just click the big, bold B than to go through several steps to create a style called "strong emphasis" and ultimately... click a big, bold B to define what that style means. This doesn't mean that styles aren't much more powerful. It doesn't mean the facility isn't important: try finding any large business that doesn't require some Marketing-designed official company template be used to give a consistent look to their slideshows! But it does illustrate that a typical user today prefers simple tools they can readily understand to more powerful tools they have to configure first. It would be a brave company that designed a user interface against that principle. They might become a spectacularly successful company if they got it right, but they'd still be brave to try!
For what it's worth, I think the hardest point of designing a user interface for any widely used application is balancing the learning curve with ease of use for power users. You can make everything use hand-holding wizards and one-step commands, but that gets in the way of experienced users who already know they want a certain combination of effects without spelling them out every time. On the other hand, even if something is easy to use and much more powerful once it's been configured the first time, requiring that configuration step can be a big hurdle for the novice. I've always thought the ideal approach would be to have the powerful underlying model based on configuration, and then to make the hand-holding stuff for new users just a front-end that leads naturally into doing everything the more powerful way, so as users gain experience it becomes the normal way to do things without there being some specific point when you switch from "novice" to "power user". Of course, I don't get to design user interfaces for products that make half my company's entire revenue stream and have hundreds of millions of users so I'm a little unqualified to comment on the practicality of this theory.:-)
Why why why why is it automatically assumed that encryption by non-government entities is in actual fact an attempt to cover up illegal activity?
Probably because statistically, a lot of it is to cover up illegal activity.
That's not an excuse for making the assumption, and it's certainly not an excuse for treating is as evidence in any sort of legal action, but it does explain the mindset.
One solution to the fact that a relatively high proportion of encrypted Internet traffic is illegal is to ban the encryption. Of course, another is to encrypt everything, as it long should have been anyway to prevent anyone other than the sender and the intended recipient accessing private data.
That's because you read too many posts on Slashdot by people who think it's more convincing to play silly word games than to make a real argument.
There are plenty of legitimate reasons to attack abusive DRM schemes like this, and plenty of cogent arguments you can make to do so. We don't need the cheap parlour tricks.
My country (the UK) isn't at the top of the slippery slope, it's falling over at the bottom. We have the legal basis right now for detention without trial, suppression of peaceful protest, arbitrary restrictions on movement (under several different laws now, actually), criminalisation based on what books you read or Internet sites you visit, arbitrary stop and search by the police, and a database state with mandatory ID cards to help enforce it all.
The only way they get away with it is because the abuses have been relatively subtle so far. A few people under house arrest, if you conveniently label them terrorist suspects, don't cause too many ruffled feathers, you see. Protests outside Parliament? Well, no-one likes them, they're just an eyesore we're better off without. A few incorrect fines because ANPR caught "your" car in a charging zone hundreds of miles from where you really were at the time? Well, some people won't even bother going to court (since it costs more than paying the incorrect fine) and the others will probably be happy just to get their money back. A few people evicted from the family homes they've lived in for decades to build a new motorway over the top? It's for the Greater Good, you see. One elderly party veteran and Holocaust survivor thrown out of a conference for a one-word heckle of a minister who took us illegally to war, and then refused readmission using anti-terrorism legislation? The law wasn't meant to be used that way. What, it was anyway? Bad policemen, no biscuit! A dead government worker here, an immigrant shot seven times in the head there, it doesn't matter, we're winning the fight to protect people's freedoms against the evil terrorists, especially the most important freedom, which is the freedom to be safe from harm, you see!
Well, I've been looking for some worthy causes to support recently. I think I'll make one of my next donations to Liberty.
(Disclaimer: No doubt many people involved in these things are doing them with the best of intentions, and yes, of course many of the people affected are nasty pieces of work. But that's not the point — at least, not if you're one of the innocent people who winds up on the wrong end of the law because of a mistake.)
Funnily enough, I was on holiday in Italy earlier this year, too. When we arrived, we too were waved right through passport control, and the only visible security in evidence was a guy with a sniffer dog as people came off the plane. We collected our baggage and headed right out into the sunshine. Welcome to Italy!
When we got back to Stansted, we stopped at passport control. Didn't have much choice, actually, since the queues were half an hour long. Everyone had to wait behind a line while the person at the front stepped forward to get their passport checked with some funky machine. Several heavily armed police officers were patrolling prominently just the other side of the inspection points. Welcome to England!
Personally, I think the Italians have it right. It's all security theatre: the biggest target at Stansted, by far, isn't the queues when people are landing, it's the entirely unsecured waiting area before people even check in. I'd rather live free and enjoy life than spend it all worrying about an over-hyped threat from which none of us can ever be completely protected anyway.
Not everything can be addressed through technology. This is such a case. Note that the original error was with a human being that chose to be duped by a phishing expedition.
True, but this story appears to have started with an employee of an outside service, salesforce.com, succumbing to phishing.
While you can't entirely beat sociological threats through technological defences, this case doesn't exactly support the standard software-as-a-service provider's argument that by outsourcing your data handling to them, you are avoiding the complexity and problems of doing it yourself. What next, confidential planning documents from a company using one of the web-based office suites get leaked after the office suite business gets tricked? There is a lesson to be learned here.
In that case they don't even need this ruling, communications between individuals outside of the US may be legally intercepted by the US Government at any time the Federal Government believes there is national interest at stake and has the ability to do so.
Sure they can... under US law.
<sigh> And people wonder why the rest of the world thinks having things like the Internet so heavily tied to US interests is a bad idea.
Also, you can't reasonably expect any privacy in email unless you encrypt its contents.
Technically, that's true, of course. But it most certainly is reasonable to expect my government not to waste time and money passing and using laws to read them without good reason, all in my name and funded by my taxes.
That's an interesting question. The spirit of trademark law is essentially to prevent someone passing off their products or services as those of another, and thus potentially taking advantage of the other's good reputation for their own benefit or damaging the other's reputation by supplying substandard products or services. Therefore if you have a blog that presents useful content, one would like to think that the identity of the blog would be protected against someone else passing themselves off as you. Whether this would apply under trademark law if you aren't currently commercialising the blog in any way and haven't registered the mark is one for the lawyers, though. I find it all too easy to believe that the ruling will wind up saying it's protected if you have a little AdSense ad somewhere that makes it "commercial", yet unprotected if you're still ramping up the readership but don't yet do anything with it money-wise, or something equally contrary to the spirit of the law.
There's a (longish) explanation of how it works in this pdf (1 MB) - the quadratic formula isn't there, but there's some examples on pp. 5 and 6. As far as I can tell, all of it is set with the Cambria (Math) font. To me it looks better on screen than the CM font, but I haven't compared them on paper.
Ouch.
I scanned the first part of that paper, and I see four obvious conclusions:
A screen font looks better on screen than a font designed for printing at fairly high resolution.
A lot of the information in the Unicode Nearly Plain Text Encoding is implicit, which gives the conciseness of notation but requires arbitrary hacks whenever what it assumes isn't actually what you want (as when you want to display the parentheses in the fraction ((a + c))/d, or you have to disambiguate bracketing operators that open and close with the same symbol such as |a|b-c|d|).
The typographic output quality is horrible in some common cases, particularly those that are or should be set with larger symbols (fractions, summation, cases, parentheses around compound expressions, etc.).
The person who wrote the paper doesn't like fair comparisons (as with the use of things like the display math markers, $$, in the TeX examples, while the equations for the Unicode Nearly Plain Text Encoding example are just magically set displayed.
The idea might be a laudable one — I'm no more a fan of TeX's verbosity than the next guy — but this isn't a fair comparison, and the output quality isn't up to professional standards yet. Meanwhile, if you adopted something like XeTeX that brings Unicode and OpenType fonts into the TeX world, you could still have all the readability advantages of writing a gamma directly instead of \gamma and you could still use a nice screen-friendly font, thus gaining the major benefits of the alternative shown in the paper, but without sacrificing the notational precision and output quality of TeX.
I think you have to keep in mind that MathML is intended to be a more general mark-up than what you get in TeX or your typical word processor's equation editor. For example, the ⁢ entity in MathML means the presentation is unambiguous and can be parsed in different ways, perhaps even spoken by a screen reader, and has no equivalent in the other notations under discussion here. MathML is verbose, and certainly not friendly to human writers, but it was never intended to be a replacement for TeX-style mark-up.
You asked why some people don't like the CM fonts, but then described one of the major problems: on-screen use. The only way to get text set in CM fonts to be reasonably legible, never mind readable, on a computer screen is to zoom right in. By doing that, you typically make the text column much wider than is comfortable for the human eye to track at a normal reading distance from the monitor, and thus make it unnecessarily difficult to read the whole body of text even if you can make out the individual characters clearly.
Compare reading an article set for on-screen reading, which would be about half the width of the monitor I'm using now and displayed in something like Georgia, Verdana, or one of the new MS fonts, with a typical zoomed PDF of an article set in CM and displayed filling the screen, and the screen font wins by several miles on both legibility and readability.
Another pet peeve of mine: why is the spacing so huge? Can someone with better knowledge of how these things work explain? Is it to make room for some of the characters?
I haven't managed to download the beta yet. Can you clarify what you think is funny about the spacing?
In terms of inter-character spacing, an OpenType font would normally feature kerning so the fit was natural at the design size when setting sentence-style material. If you're setting in all-caps (not generally a good idea, but sometimes useful for effect) then in many fonts it looks better if you would increase the tracking slightly. Setting intricate mathematical expressions is very difficult to do well just through font features, though: even TeX provides numerous small spacing adjustment commands for the many occasions when things don't quite come out right using the mechanical defaults.
In terms of inter-line spacing, most OpenType fonts include enough leading that if you set some text in a simple application that just spaces the lines at the font size, the text will look OK and you won't get clashes between ascenders and descenders or similar silliness. More sophisticated applications should be able to get the various spacing metrics from the font file and take these into account when setting the leading explicitly, e.g., 12pt text with an additional 3pt of leading.
For a mathematical context, slightly wider line-spacing is often used compared to something like a novel where you're just setting paragraphs of plain text. This allows space for notations like super- and subscripts and symbols such as sums and integrals to be set in-line without clashing with the lines above or below. Perhaps if you're worried about the line-spacing in the STIX fonts, the default is wider than you're used to for this reason? One would expect any decent word processor or typesetting package to take this into account if you're specifying more precise typographical requirements, though, so if you're setting in say 11/15pt it shouldn't look much different whether you're using an OpenType font with wider default line-spacing or not.
Windows is a bit funny when it comes to font smoothing. The standard option has a font size floor, below which nothing is smoothed. IMHO, this is actually rather large. ClearType seems to try to smooth everything regardless, which leads to the bizarre situation that I actually prefer to have ClearType on even on a CRT, because text at moderate sizes looks much better antialiased even if the subpixel effects sometimes cause artifacts because they were designed for TFT screens.
If anything can do it, it'll be an initiative something like the STIX work.
In any case, Computer Modern is far from everyone's taste. Knuth did a great job designing a highly legible font that could both typeset mathematics elegantly and survive the scanning, photocopying and other abuse scientific papers tend to suffer. However, notwithstanding Knuth's personal preferences, aesthetically the Computer Modern set leaves a lot to be desired. Many people prefer a different style on paper, and on screen the lightness of the CM set is pretty horrible, as anyone who's tried to read a long PDF of a paper set using TeX can testify.
It's a shame that in a world where OpenType and Unicode are now commonplace, and where many professional fonts now come with glyphs for numerous different alphabets and numerous carefully tuned typographical features, it isn't yet common to supply matching glyphs for say the top 100 scientific symbols. I guess the market is just too specialised, and the current dominance of the TeX family means there's little commercial incentive for others to produce high quality scientific fonts. In that respect, having a high-quality, science-friendly font available for use with things like web pages surely must be a good thing. (Monospace fonts useful for typesetting computer programs currently suffer a similar lack of support, probably for the same reason.)
I'm sorry, you must have been on another planet. The entire British media was full of stories about the controversy of Gordon Brown taking over as Prime Minister based only on the votes of Labour Party members and their political allies. Then it was full of stories about how undemocratic it was to have what they termed a "coronation", where Brown didn't even have to face any other candidates from within his own party.
And remember, the Labour MPs who currently hold an absolute majority in Parliament — including Brown himself — were all elected based on an explicit promise that Tony Blair would serve a full third term as PM, which in turn was motivated by the huge amount of concern that by voting Labour people might be voting for Gordon Brown.
There's just no way you can credibly argue that Brown taking over as PM was even close to democratic or representative of the will of the people. Our political system is pretty crooked at the best of times, but this is a new low. But now I've gone and fed a troll by reading Slashdot too late at night. Damn.
Slashdot Mirror
The significant fact to my thesis was that they will delete your information.
No, they won't. They will hide it, and can bring it back on demand.
If you don't get the difference, then you need to talk to a few more of those "most pedantic privacy advocates" and find out how many all too serious problems happen to ordinary people every year because "deleted" information such as personal details or account numbers wasn't actually deleted at all. Perhaps you've heard of identity theft? Stalking? Credit card fraud?
Your friends can post pictures of you with your name in any media ranging from Flickr to a bathroom stall.
Actually, systematically collecting a database of personally identifiable information about anyone without their consent is illegal in some places, and comes with significant limitations on what you're allowed to do in many more. The fact that it's legal at all is a problem, as is the fact that not everywhere gives people a right to force people storing data about them to delete it — properly — if there isn't a good reason to keep it around.
Society currently fails to recognise the dangers inherent in allowing this sort of behaviour. It will, before long, when enough people get hurt. It's just a shame that until everyone understands the idea that just because we can do something it doesn't mean we should, a lot of people are going to get hurt needlessly.
In what way was my post incorrect? Accessing any of the content hosted on Facebook so you can see how it works requires a user account, or at least it did at the time I signed up. There is no way one could, to give the obvious example, conduct a simple experiment to see how much information they really collect without signing up first (or working with someone else who did so).
I'm sorry, but every single thing you wrote in your first paragraph is objectively, factually incorrect — other than the fact that you're not so sure about some of it, but that just makes you wrong.
How that got a (+1, Informative) is beyond me.
You didn't do your research, and want to complain about it after the fact while taking no personal responsibility.
Ah, another genius who not only has time to read all the detailed agreements he ever sees and understand them, but is also blessed with prescience that allows him to interpret the intentions of other parties ahead of time.
So how do you propose I should do some research on the nature of the web site, given that to gain any access at all requires creating an account, and that it is the concessions made when creating the account that is in question here?
You can only take responsibility for something you have some reasonable control over. There was no realistic way to know what would happen in this case.
I think the view you present is perhaps too simple. There is no one check box that says a patent is "legal" or not. There is a question of whether a patent has been granted by a patent awarding body. There is a question of how widely such a patent is recognised. There is a question of whether infringing claims in an awarded patent is legally actionable in any given jurisdiction.
It is perfectly rational to argue that a patent was awarded lawfully by a recognised body, regardless of whether any legal action could successfully be brought against someone infringing it in a given jurisdiction at the time. You could also make a rational case that if a patent was properly awarded by the relevant body, no previous test had found it unenforceable in court, and infringement is actionable, then there has been no change in the legal situation.
Now, I'm not offering any view on the morality of taking such a position, nor claiming that the law would necessarily support it, nor necessarily disagreeing with the general ethical principle that retroactive laws are a bad idea. I'm just saying that a good lawyer could surely build a reasonable case, and it's nowhere near as safe or obvious to ignore dubious software patents as you seem to be suggesting.
I find the Facebook privacy stories frustrating because they seem to always ignore one thing - in almost every case the Facebook user decides how much information to make public, to whom, and which applications to install.
Sorry, but that's just not true.
Facebook are quite happy to collect information on anyone who has ever been a user, including identifying them in photographs, even if they closed their account immediately after discovering that the site is one big invasion of privacy. Facebook offer no mechanism for ex-users to permanently delete such information, nor to prevent others continuing to provide it after a user cancels their account (despite the fact that this is almost certainly illegal in many jurisdictions).
So what next? Anyone whose friends group has ever mentioned a product on someone's wall is consenting to to any image of them tagged in a photograph by someone else without their knowledge being used in advertising? Anyone who once mentioned something privately to their friends in order to criticise it gets their face used to promote that thing to the world?
There is just no excuse for this. It's exactly why I am the guy who quit Facebook almost as soon as I'd joined it. Facebook, like Google, is one of the biggest dangers to modern society. Society just hasn't realised it yet, and lets them get away with stuff because they present the appearance of a useful service. Pandora's box ought to be required reading in schools.
Why is CMYK so important, don't the new, modern presses use RGB natively now?
It's not that simple. Sure, many places can convert RGB material if that's what you supply, but it doesn't have the same reliability and flexibility as CMYK for professional quality work. This is pretty obvious if you think about it, given both that CMYK has an extra "axis" and that CMYK is what will typically be used for the physical printing itself.
On a different but still relevant note, it's also considerably cheaper to use just one or two spot colours rather than a full colour print if you don't need the whole range. However, if you're using specific choices from something like a Pantone swatch, you really need some software support to use those effectively. As I understand it, the GIMP doesn't do this yet, either.
For the record, I don't work in media, but I have done publicity work for large organisations involving dealing with external print shops. I guess you might call my perspective a "semi-professional" point of view.
Your point about software patents outside the US is well taken, and indeed even in the US I would think there was a huge amount of prior art to invalidate all but the most specific claims.
However, I'm afraid things aren't as simple elsewhere as your retroactive law argument suggests. This is why certain companies are trying to get what are effectively software patents onto the books in more permissive jurisdictions within Europe (I seem to recall Ireland being mentioned as a favourite target for this, though I'm not sure what makes it more appealing). The law doesn't currently say patents can't be awarded in category X, it just says which ones can be. However, if you can get something onto the books that wouldn't actually be enforceable under the patent laws in European countries today, but big business succeeds in its repeated attempts to legitimise software patents somewhere down the line, then it won't be a retroactive measure because the patent was approved earlier. As with many things involving big business and the law, it's not necessarily the spirit that counts, but how you play the game and what technicalities you know about the rules. :-(
While I can certainly understand why you might want to work that way yourself, I think there might be a few practical difficulties with applying your reasoning more generally.
For one thing, you mentioned effectively training the user on demand when they need to do new things, but both before-the-fact training and after-the-fact on-line help or support calls work much better if the context is consistent. I notice that the Microsoft Office team — who, for all their sins, are pretty careful about their user interface and actively collect feedback — have moved in exactly the opposite direction. The magic hiding menus feature was pretty much universally slammed, and with Office 2007 they've gone for something much less customisable than before. And despite the bitching on Slashdot because they're Microsoft and they changed something, I've yet to meet someone in the real world who didn't say they preferred the new version after a little time getting used to it.
A related point is that for all the UI customisability in many modern applications, almost no-one actually uses it. Simple things like setting up styles and templates in word processors or presentation packages are ignored in favour of ad-hoc formatting. Personally, I think this has a lot to do with the fact that UIs make it easier to just click the big, bold B than to go through several steps to create a style called "strong emphasis" and ultimately... click a big, bold B to define what that style means. This doesn't mean that styles aren't much more powerful. It doesn't mean the facility isn't important: try finding any large business that doesn't require some Marketing-designed official company template be used to give a consistent look to their slideshows! But it does illustrate that a typical user today prefers simple tools they can readily understand to more powerful tools they have to configure first. It would be a brave company that designed a user interface against that principle. They might become a spectacularly successful company if they got it right, but they'd still be brave to try!
For what it's worth, I think the hardest point of designing a user interface for any widely used application is balancing the learning curve with ease of use for power users. You can make everything use hand-holding wizards and one-step commands, but that gets in the way of experienced users who already know they want a certain combination of effects without spelling them out every time. On the other hand, even if something is easy to use and much more powerful once it's been configured the first time, requiring that configuration step can be a big hurdle for the novice. I've always thought the ideal approach would be to have the powerful underlying model based on configuration, and then to make the hand-holding stuff for new users just a front-end that leads naturally into doing everything the more powerful way, so as users gain experience it becomes the normal way to do things without there being some specific point when you switch from "novice" to "power user". Of course, I don't get to design user interfaces for products that make half my company's entire revenue stream and have hundreds of millions of users so I'm a little unqualified to comment on the practicality of this theory. :-)
Why why why why is it automatically assumed that encryption by non-government entities is in actual fact an attempt to cover up illegal activity?
Probably because statistically, a lot of it is to cover up illegal activity.
That's not an excuse for making the assumption, and it's certainly not an excuse for treating is as evidence in any sort of legal action, but it does explain the mindset.
One solution to the fact that a relatively high proportion of encrypted Internet traffic is illegal is to ban the encryption. Of course, another is to encrypt everything, as it long should have been anyway to prevent anyone other than the sender and the intended recipient accessing private data.
That's because you read too many posts on Slashdot by people who think it's more convincing to play silly word games than to make a real argument.
There are plenty of legitimate reasons to attack abusive DRM schemes like this, and plenty of cogent arguments you can make to do so. We don't need the cheap parlour tricks.
and so it begins.....
It begins?!
My country (the UK) isn't at the top of the slippery slope, it's falling over at the bottom. We have the legal basis right now for detention without trial, suppression of peaceful protest, arbitrary restrictions on movement (under several different laws now, actually), criminalisation based on what books you read or Internet sites you visit, arbitrary stop and search by the police, and a database state with mandatory ID cards to help enforce it all.
The only way they get away with it is because the abuses have been relatively subtle so far. A few people under house arrest, if you conveniently label them terrorist suspects, don't cause too many ruffled feathers, you see. Protests outside Parliament? Well, no-one likes them, they're just an eyesore we're better off without. A few incorrect fines because ANPR caught "your" car in a charging zone hundreds of miles from where you really were at the time? Well, some people won't even bother going to court (since it costs more than paying the incorrect fine) and the others will probably be happy just to get their money back. A few people evicted from the family homes they've lived in for decades to build a new motorway over the top? It's for the Greater Good, you see. One elderly party veteran and Holocaust survivor thrown out of a conference for a one-word heckle of a minister who took us illegally to war, and then refused readmission using anti-terrorism legislation? The law wasn't meant to be used that way. What, it was anyway? Bad policemen, no biscuit! A dead government worker here, an immigrant shot seven times in the head there, it doesn't matter, we're winning the fight to protect people's freedoms against the evil terrorists, especially the most important freedom, which is the freedom to be safe from harm, you see!
Well, I've been looking for some worthy causes to support recently. I think I'll make one of my next donations to Liberty.
(Disclaimer: No doubt many people involved in these things are doing them with the best of intentions, and yes, of course many of the people affected are nasty pieces of work. But that's not the point — at least, not if you're one of the innocent people who winds up on the wrong end of the law because of a mistake.)
Funnily enough, I was on holiday in Italy earlier this year, too. When we arrived, we too were waved right through passport control, and the only visible security in evidence was a guy with a sniffer dog as people came off the plane. We collected our baggage and headed right out into the sunshine. Welcome to Italy!
When we got back to Stansted, we stopped at passport control. Didn't have much choice, actually, since the queues were half an hour long. Everyone had to wait behind a line while the person at the front stepped forward to get their passport checked with some funky machine. Several heavily armed police officers were patrolling prominently just the other side of the inspection points. Welcome to England!
Personally, I think the Italians have it right. It's all security theatre: the biggest target at Stansted, by far, isn't the queues when people are landing, it's the entirely unsecured waiting area before people even check in. I'd rather live free and enjoy life than spend it all worrying about an over-hyped threat from which none of us can ever be completely protected anyway.
Not everything can be addressed through technology. This is such a case. Note that the original error was with a human being that chose to be duped by a phishing expedition.
True, but this story appears to have started with an employee of an outside service, salesforce.com, succumbing to phishing.
While you can't entirely beat sociological threats through technological defences, this case doesn't exactly support the standard software-as-a-service provider's argument that by outsourcing your data handling to them, you are avoiding the complexity and problems of doing it yourself. What next, confidential planning documents from a company using one of the web-based office suites get leaked after the office suite business gets tricked? There is a lesson to be learned here.
In that case they don't even need this ruling, communications between individuals outside of the US may be legally intercepted by the US Government at any time the Federal Government believes there is national interest at stake and has the ability to do so.
Sure they can... under US law.
<sigh> And people wonder why the rest of the world thinks having things like the Internet so heavily tied to US interests is a bad idea.
Also, you can't reasonably expect any privacy in email unless you encrypt its contents.
Technically, that's true, of course. But it most certainly is reasonable to expect my government not to waste time and money passing and using laws to read them without good reason, all in my name and funded by my taxes.
That's an interesting question. The spirit of trademark law is essentially to prevent someone passing off their products or services as those of another, and thus potentially taking advantage of the other's good reputation for their own benefit or damaging the other's reputation by supplying substandard products or services. Therefore if you have a blog that presents useful content, one would like to think that the identity of the blog would be protected against someone else passing themselves off as you. Whether this would apply under trademark law if you aren't currently commercialising the blog in any way and haven't registered the mark is one for the lawyers, though. I find it all too easy to believe that the ruling will wind up saying it's protected if you have a little AdSense ad somewhere that makes it "commercial", yet unprotected if you're still ramping up the readership but don't yet do anything with it money-wise, or something equally contrary to the spirit of the law.
There's a (longish) explanation of how it works in this pdf (1 MB) - the quadratic formula isn't there, but there's some examples on pp. 5 and 6. As far as I can tell, all of it is set with the Cambria (Math) font. To me it looks better on screen than the CM font, but I haven't compared them on paper.
Ouch.
I scanned the first part of that paper, and I see four obvious conclusions:
The idea might be a laudable one — I'm no more a fan of TeX's verbosity than the next guy — but this isn't a fair comparison, and the output quality isn't up to professional standards yet. Meanwhile, if you adopted something like XeTeX that brings Unicode and OpenType fonts into the TeX world, you could still have all the readability advantages of writing a gamma directly instead of \gamma and you could still use a nice screen-friendly font, thus gaining the major benefits of the alternative shown in the paper, but without sacrificing the notational precision and output quality of TeX.
I think you have to keep in mind that MathML is intended to be a more general mark-up than what you get in TeX or your typical word processor's equation editor. For example, the ⁢ entity in MathML means the presentation is unambiguous and can be parsed in different ways, perhaps even spoken by a screen reader, and has no equivalent in the other notations under discussion here. MathML is verbose, and certainly not friendly to human writers, but it was never intended to be a replacement for TeX-style mark-up.
You asked why some people don't like the CM fonts, but then described one of the major problems: on-screen use. The only way to get text set in CM fonts to be reasonably legible, never mind readable, on a computer screen is to zoom right in. By doing that, you typically make the text column much wider than is comfortable for the human eye to track at a normal reading distance from the monitor, and thus make it unnecessarily difficult to read the whole body of text even if you can make out the individual characters clearly.
Compare reading an article set for on-screen reading, which would be about half the width of the monitor I'm using now and displayed in something like Georgia, Verdana, or one of the new MS fonts, with a typical zoomed PDF of an article set in CM and displayed filling the screen, and the screen font wins by several miles on both legibility and readability.
Another pet peeve of mine: why is the spacing so huge? Can someone with better knowledge of how these things work explain? Is it to make room for some of the characters?
I haven't managed to download the beta yet. Can you clarify what you think is funny about the spacing?
In terms of inter-character spacing, an OpenType font would normally feature kerning so the fit was natural at the design size when setting sentence-style material. If you're setting in all-caps (not generally a good idea, but sometimes useful for effect) then in many fonts it looks better if you would increase the tracking slightly. Setting intricate mathematical expressions is very difficult to do well just through font features, though: even TeX provides numerous small spacing adjustment commands for the many occasions when things don't quite come out right using the mechanical defaults.
In terms of inter-line spacing, most OpenType fonts include enough leading that if you set some text in a simple application that just spaces the lines at the font size, the text will look OK and you won't get clashes between ascenders and descenders or similar silliness. More sophisticated applications should be able to get the various spacing metrics from the font file and take these into account when setting the leading explicitly, e.g., 12pt text with an additional 3pt of leading.
For a mathematical context, slightly wider line-spacing is often used compared to something like a novel where you're just setting paragraphs of plain text. This allows space for notations like super- and subscripts and symbols such as sums and integrals to be set in-line without clashing with the lines above or below. Perhaps if you're worried about the line-spacing in the STIX fonts, the default is wider than you're used to for this reason? One would expect any decent word processor or typesetting package to take this into account if you're specifying more precise typographical requirements, though, so if you're setting in say 11/15pt it shouldn't look much different whether you're using an OpenType font with wider default line-spacing or not.
Windows is a bit funny when it comes to font smoothing. The standard option has a font size floor, below which nothing is smoothed. IMHO, this is actually rather large. ClearType seems to try to smooth everything regardless, which leads to the bizarre situation that I actually prefer to have ClearType on even on a CRT, because text at moderate sizes looks much better antialiased even if the subpixel effects sometimes cause artifacts because they were designed for TFT screens.
If anything can do it, it'll be an initiative something like the STIX work.
In any case, Computer Modern is far from everyone's taste. Knuth did a great job designing a highly legible font that could both typeset mathematics elegantly and survive the scanning, photocopying and other abuse scientific papers tend to suffer. However, notwithstanding Knuth's personal preferences, aesthetically the Computer Modern set leaves a lot to be desired. Many people prefer a different style on paper, and on screen the lightness of the CM set is pretty horrible, as anyone who's tried to read a long PDF of a paper set using TeX can testify.
It's a shame that in a world where OpenType and Unicode are now commonplace, and where many professional fonts now come with glyphs for numerous different alphabets and numerous carefully tuned typographical features, it isn't yet common to supply matching glyphs for say the top 100 scientific symbols. I guess the market is just too specialised, and the current dominance of the TeX family means there's little commercial incentive for others to produce high quality scientific fonts. In that respect, having a high-quality, science-friendly font available for use with things like web pages surely must be a good thing. (Monospace fonts useful for typesetting computer programs currently suffer a similar lack of support, probably for the same reason.)
I'm sorry, you must have been on another planet. The entire British media was full of stories about the controversy of Gordon Brown taking over as Prime Minister based only on the votes of Labour Party members and their political allies. Then it was full of stories about how undemocratic it was to have what they termed a "coronation", where Brown didn't even have to face any other candidates from within his own party.
And remember, the Labour MPs who currently hold an absolute majority in Parliament — including Brown himself — were all elected based on an explicit promise that Tony Blair would serve a full third term as PM, which in turn was motivated by the huge amount of concern that by voting Labour people might be voting for Gordon Brown.
There's just no way you can credibly argue that Brown taking over as PM was even close to democratic or representative of the will of the people. Our political system is pretty crooked at the best of times, but this is a new low. But now I've gone and fed a troll by reading Slashdot too late at night. Damn.