And mine are on the street if they learn PHP at any point... Although, actually, children learn to talk before they learn to draw, so why should mastering computer language come 12 years after mastering point and click? Or, conversely, why are GUIs so shallow?
The GUIs we have now generally have the grammatical sophistication of a 2 year-old: "Want THAT, Don't want THAT, move THAT, and THAT, and THAT..." Seeing Kay's name go past again makes me nostalgic for an age when interface designers dreamed bold dreams, when object-oriented meant more than another way of managing libraries, when we really thought that GUIs could become languages rather than cave paintings revisited. I remember playing with GUI code on a Symbolics 15 years ago that still makes any modern GUI seem incredibly limited, and this despite having 100 times the processing power we had then.
I think we settled for a parody of the original dream.
LTSP is great, but I don't think it will do the 56k modem thing. It's basically X11 and a few other standard Linux services linked together in a rather clever way, so the client-server networks overheads are the same as for X11, plus running XFS over tcp, plus serving files... with ten machines, a 10mbit network boils, our current 100mbit one gets congested on occasions, and I'm thinking of getting a gigabyte backbone from the hub to the server in the near future.
What you can do is run rdesktop, mentioned below, from an LTSP terminal. So if your terminals are in the same place, you could have a local network with a modest Linux server running the diskless terminals, and those terminals connecting to W2K using rdekstop and Terminal Server. You still need TS licences, but they are a lot cheaper than Citrix licences, especially for education. I've tried it briefly, and it worked very well: just waiting for the other half of my Windows server to turn up in the post to try it for real...
I fear that we are heading into 'real engineers don't eat quiche' territory here...
Surely the point is that the guy is talking about creating stuff from scratch, and most programmers and most engineers don't do that most of the time. If you are building bridges, most of the time you take an existing design, tweak it a bit, and ship it. And a lot of programming is like that too.
Now I suspect that engineers who are working on a fundamentally new type of structure look a lot like programmers who are writing some radical new program from scratch. The nearest equivalent to what the latter feels like to me is composing polyphonic music scores, which most people would consider to be art, but which you could equally consider to be applied math.
And if you haven't met more than three female programmers, all I can say is that you need to get out more:-)
I can't quite believe I'm reading most of this thread. I thought the US was for opening markets for free trade. I thought us Europeans were the protectionist ones. If we want to scrap GATT and go back to building warring trade blocs, plenty of European governments will be happy to impose punitive tarifs on American software, cars, electronics products and so on. And the countries who export the most are the ones who will lose the most by such a move.
Ideologically, France, for example, would like nothing more than to ban American software - I saw an article in the Figaro the other week about imposing import levies on foreign computer games. But, at present, the consumers' wish to buy American products carries the day. And you guys want to hand the government a loaded gun pointed at your own heads?!?!
Wouldn't another option be to ask the postwar Iraqi regime what system they want? I mean, this war is about giving the Iraqis real freedom, not replacing one dictatorship with another, right?
On paper, thin clients have a lot going for them. My cybercafe has 10 of them, and, in general, they are a big win over peer networking. The problem is that the cheapest way to get a thin client is to build half a PC, rather than any of the more elegant bespoke solutions. If Sun had a cheap terminal that let a roomful of people pretend they had a Sun workstation of their own, I can see a few people being interested...
the well-publicised 'out of the box' antialiased fonts
I upgraded from 7.3 to 8.0 last week, which was generally considered to be a Good Thing by our customers, but one of the bits that doesn't seem to work very well is the fonts
For example, if I type an e acute (é, but not using a glyph) in HTML, with a latin1 charset specified in the metatags, it displays as 2 garbled characters with Phoenix, but displays correctly with Konqueror. On the other hand, arabic fonts work with Phoenix but display as little squares with Konqueror. Similar odd problems with kmail.
We added a couple of true type libraries to xfs, which might have something to do with it, but kmail is using the standard kmail font. Is anyone else having these sort of problems?
"It is not a desktop, but works on a big server platform," he said.
He appears to be talking about some sort of thin client, which is certainly potentially different to what MS is doing. Whether it is actually any thinner than a PC running terminal-type software, and whether Sun can do any better than Oracle in making thin clients take off beyond a few specific niches remains to be seen.
Apart from the interesting idea of hinduism as non-prosetylising (maybe true in the West, but plenty of Indians have died trying to leave hindiuism), the hype surrounding Linux is astounding, and this article is a typical example. What is the combined desktop market share of the 40% of developers working on Linux? I use Linux every day for my work, but I can see some very good reasons why people would want to stick to Windows on their desktop, at least for now. It is bad for anyone, religious or otherwise, to believe their own propaganda.
I don't know if the story is true, but, if it isn't, it should be. You are wrong about the US Army though: the bridge would still be standing and the village nearby would right now be passing Pluto:-)
The houses that have lasted a hundred years are the good ones.
Yes, a lot of the examples on this thread basically prove that rich people can afford to build houses that last. But it doesn't always work that way.
Our house in the Luberon (SE France) is probably about 600 years old, and was probably built by a group of nomads who settled in our area and manifestly didn't know much about building. So they started by digging half the rooms out of the (very soft) bedrock, added a few barrel vaults made from what they had dug out, made all the walls 3 foot thick and kept the distance between walls to less than 12ft. The walls themselves are two piles of soft stone held together with lime mortar, with the gap filled up with whatever they could find (including a lot of straw AFAICS.) It's the weight of the infill rubble that keeps the vaults strong.
Now the advantage of having a house that is basically a slightly organised pile of rubble is that, in entropy terms, there just isn't very far for it to degrade. When we bought it there was a huge crack down one wall, and we just filled it up with some more boulders and lime mortar.
All this in a low grade earthqake zone, where the French army tried to burn most of the villages to the ground at one point, but the only effect of this was to make the roofs cave in and make the walls black (as you see when you start drilling holes).
On a related note, the first suspension bridge in the world, in Bristol, UK, is also one of the most stable. Why? Because, like most things that Brunel built, it is overengineered by several orders of magnitude. If he had had a beowulf cluster to do his simulations on, it would be a lot lighter, and would probably have fallen into the Avon Gorge some time ago:-)
Well, yes, but since my business is letting people look at the sites they want to, whether I think they suck or not, that doesn't really help. I suppose I could try starting the world's first strict xhtml4-only cybercafe, but I suspect that the 3 people in France who would be impressed have Internet connections of their own:-) For the rest of the population, the Internet is a consumer product, and they expect it to work. They don't care how it works, they do care if they can't check their bank balance or whatever.
List of bugs?
I don't have my own, there are various comparisons out there on the net. The last one I found was a real classic, a website where mouseover made the entire paragraph split onto 2 lines. Absolutely nothing in the CSS to justify this. We eventually decided that Mozilla was changing the size of the paragraph element depending on its background colour. The background colour attribute has now been removed, which fixed the problem, so unfortunately I can't demo it, but I showed it to a dozen or so people before they fixed it and everyone agreed that it was quite surreal. I think someone filed it with bugzilla.
Last night I started reading a book by Eric Meyer, and one of the points he makes early on, to my surprise, is that it's a bad move to try to do everything with CSS because browser support is so patchy. And he's a Standards Evangelist for Netscape... so that leaves all of us producing more or less sucky markup depending on how we evaluate the browser mix likely to look at our site this week.
Finally, I don't think the 'MS incompatible to take control of Internet' thing holds water in this particular case. Meyer points out that there have been changes between IE5 and IE6 which move IE closer to W3C standards, and which thus mean that the only way to make a page display properly on both versions of IE is to use the same workaround that was always needed to make pages viewable by IE5 and Mozilla. If that's an example of MS trying to push out other browsers, they need a new strategist:-)
Unless you want to do Internet banking, or reliable webchat, or be able to see any site that has only been tested with IE, or which blocks access from any other browser. I use Mozilla in my cybercafe, and there are a significant number of sites that customers want to see that just don't work. I cancelled my own online banking contract because I couldn't get into the site.
And before anyone says that I should change bank, or tell my customers not to use non w3c-compliant websites, or something, can we try to stay in the real world? Most people don't plan their entire life around their choice of browser.
Also, from a webmaster's perspective, I find Mozilla CSS non-compliance anything but easy to work around. Sure, the big things are right, unlike some well-documented areas of IE, but working around a couple of big deviations is a lot easier than working around thousands of small ones. I frequently spend time trying to get a website to render properly with Mozilla, only to find that it works fine with IE, Opera and Konqueror.
All being well, we install IE in the cybercafe next week.
So just to get our story straight here, AOL sending out lots of unwanted CDs is bad, but the Swiss government sending out lots of CDs is good?
Actually, given the Swiss government's rabid stance on recycling (at one point the genevois were smuggling bags of garbage into France to avoid paying garbage tax), this whole project could be considered to be a form of stealth tax, ie they send you the CD, and if you don't dispose of it properly they fine you...
They aren't the only ones who are insane, although the fact that they also have all that firepower is a little worrying.
I remember going to a particularly paranoid seminar on security, and being told about a group of European missionaries who were exchanging pgp-encrypted emails in some African country, and the secret police asked them why they were doing this, which proved how important it is to encrypt your messages. I thought it proved what a dumb idea it is: anyone screening messages is going to pick out the encrypted ones as potentially interesting, and, even if they can't break the encryption, they can always break your door down, pull out your fingernails or just confiscate your computer.
Sending plain text emails, substituting sensitive words where necessary, strikes me as a far better way to go. Although most of my emails would send even the most conscientious spy to sleep...
It's wierder than that. You are talking about an OS that has been ploughing it's own furrow on custom hardware since before Linux was a twinkle in Linus' eye, so trying to understand it in terms of standard PC hardware is not going to be very useful. As I have explained elsewhere, the whole OS is in the ROM, possibly including the fonts and some applications. So the 'analagous to BIOS' bit of the statement is just not true, IMHO.
Much as I love RISC OS and wish Castle well, this is pretty disingenious. Pretty well the entire OS is in the ROMs: with V3 you could boot the machine into a windowing environment and produce vector graphics with the equivalent of Mac Draw, complete with anti-aliased fonts, and print the result, without even inserting a floppy. I think the applications moved onto disc in v4, but there is still an awful lot more than a BIOS on the ROM.
There was a lot of talk about the industry moving this way a while back. Is it ever going to happen? Sounds eminently fair to me. The result would be far smaller bills for light users, making it easier for people to get started, and that heavy users might think a bit more about how much of this stuff they actually need. If people want to run a worldwide video distribution operation from their bedrooms, that's fine by me, but I don't really see why I should subsidise it...
Actually, I doubt it. I think you will find that the company in question has a dozen or so employees. We are not talking Microsoft here. The few people who have ever heard of them are mostly surprised that they are still in business.
Pinching the code was a naughty no-no, but, in terms of a character defence, they were standing up to Microsoft when most people on this list were still happily using DOS, and, for several years, they did a pretty good job, at least in the UK. So I hope that a negotiated settlement can be found.
the only understanding you'll get is from scientific inquiry
Depends exactly what aspect of this accident you are trying to deal with. For reducing the chances of a repeat performance, science is the way to go. For dealing with the existential questions, science is not a lot of help.
Reading this thread from the other side of the Atlantic, I would suggest that there is more going on here than a discussion of the event in reductionist terms.
Yes, 7 people died, and, yes, that's very sad. But how many people died on the DC beltway today? Why are 270 million people going to be grieving for 7 astronauts, rather than for all the other people who died today? Rationally, it makes no sense. But in terms of what the shuttle represents subjectively, expecially to Americans, it makes a lot of sense.
What this accident says to everyone watching is we are all mortal, and analysing the state of the tiles on the Shuttle doesn't begin to bring understanding of that. Death is the new taboo. Death of young people using the latest technology is as bad as it gets in our (shared) culture. That's why people want to pray, or light a candle, or punch someone, or get drunk, or something. Personally, I think praying is one of the more useful options in that list.
Ooooh, I thought the idea was that one person read the article and we all asked him questions:-)
To all intents and purposes...
I can't say I'm reassured that it would be me rather than a copy of me who stepped out of the teleporter, although the question is probably closer to theology than physics...
How do they know it's the same electron? What distinctive marks are they basing this on? If it's just one electron that disappears and another one that reappears, the applications would be rather more limited. I mean, if I step into a teleportation machine, I would quite like me rather than someone who looks just like me to step out a mile away.
Slashdot Mirror
And mine are on the street if they learn PHP at any point... Although, actually, children learn to talk before they learn to draw, so why should mastering computer language come 12 years after mastering point and click? Or, conversely, why are GUIs so shallow?
The GUIs we have now generally have the grammatical sophistication of a 2 year-old: "Want THAT, Don't want THAT, move THAT, and THAT, and THAT..." Seeing Kay's name go past again makes me nostalgic for an age when interface designers dreamed bold dreams, when object-oriented meant more than another way of managing libraries, when we really thought that GUIs could become languages rather than cave paintings revisited. I remember playing with GUI code on a Symbolics 15 years ago that still makes any modern GUI seem incredibly limited, and this despite having 100 times the processing power we had then.
I think we settled for a parody of the original dream.
LTSP is great, but I don't think it will do the 56k modem thing. It's basically X11 and a few other standard Linux services linked together in a rather clever way, so the client-server networks overheads are the same as for X11, plus running XFS over tcp, plus serving files... with ten machines, a 10mbit network boils, our current 100mbit one gets congested on occasions, and I'm thinking of getting a gigabyte backbone from the hub to the server in the near future.
What you can do is run rdesktop, mentioned below, from an LTSP terminal. So if your terminals are in the same place, you could have a local network with a modest Linux server running the diskless terminals, and those terminals connecting to W2K using rdekstop and Terminal Server. You still need TS licences, but they are a lot cheaper than Citrix licences, especially for education. I've tried it briefly, and it worked very well: just waiting for the other half of my Windows server to turn up in the post to try it for real...
I fear that we are heading into 'real engineers don't eat quiche' territory here...
Surely the point is that the guy is talking about creating stuff from scratch, and most programmers and most engineers don't do that most of the time. If you are building bridges, most of the time you take an existing design, tweak it a bit, and ship it. And a lot of programming is like that too.
Now I suspect that engineers who are working on a fundamentally new type of structure look a lot like programmers who are writing some radical new program from scratch. The nearest equivalent to what the latter feels like to me is composing polyphonic music scores, which most people would consider to be art, but which you could equally consider to be applied math.
And if you haven't met more than three female programmers, all I can say is that you need to get out more :-)
I can't quite believe I'm reading most of this thread. I thought the US was for opening markets for free trade. I thought us Europeans were the protectionist ones. If we want to scrap GATT and go back to building warring trade blocs, plenty of European governments will be happy to impose punitive tarifs on American software, cars, electronics products and so on. And the countries who export the most are the ones who will lose the most by such a move.
Ideologically, France, for example, would like nothing more than to ban American software - I saw an article in the Figaro the other week about imposing import levies on foreign computer games. But, at present, the consumers' wish to buy American products carries the day. And you guys want to hand the government a loaded gun pointed at your own heads?!?!
Wouldn't another option be to ask the postwar Iraqi regime what system they want? I mean, this war is about giving the Iraqis real freedom, not replacing one dictatorship with another, right?
On paper, thin clients have a lot going for them. My cybercafe has 10 of them, and, in general, they are a big win over peer networking. The problem is that the cheapest way to get a thin client is to build half a PC, rather than any of the more elegant bespoke solutions. If Sun had a cheap terminal that let a roomful of people pretend they had a Sun workstation of their own, I can see a few people being interested...
the well-publicised 'out of the box' antialiased fonts
I upgraded from 7.3 to 8.0 last week, which was generally considered to be a Good Thing by our customers, but one of the bits that doesn't seem to work very well is the fonts
For example, if I type an e acute (é, but not using a glyph) in HTML, with a latin1 charset specified in the metatags, it displays as 2 garbled characters with Phoenix, but displays correctly with Konqueror. On the other hand, arabic fonts work with Phoenix but display as little squares with Konqueror. Similar odd problems with kmail.
We added a couple of true type libraries to xfs, which might have something to do with it, but kmail is using the standard kmail font. Is anyone else having these sort of problems?
The clue is in the article:
He appears to be talking about some sort of thin client, which is certainly potentially different to what MS is doing. Whether it is actually any thinner than a PC running terminal-type software, and whether Sun can do any better than Oracle in making thin clients take off beyond a few specific niches remains to be seen.
Apart from the interesting idea of hinduism as non-prosetylising (maybe true in the West, but plenty of Indians have died trying to leave hindiuism), the hype surrounding Linux is astounding, and this article is a typical example. What is the combined desktop market share of the 40% of developers working on Linux? I use Linux every day for my work, but I can see some very good reasons why people would want to stick to Windows on their desktop, at least for now. It is bad for anyone, religious or otherwise, to believe their own propaganda.
I don't know if the story is true, but, if it isn't, it should be. You are wrong about the US Army though: the bridge would still be standing and the village nearby would right now be passing Pluto :-)
The houses that have lasted a hundred years are the good ones.
Yes, a lot of the examples on this thread basically prove that rich people can afford to build houses that last. But it doesn't always work that way.
Our house in the Luberon (SE France) is probably about 600 years old, and was probably built by a group of nomads who settled in our area and manifestly didn't know much about building. So they started by digging half the rooms out of the (very soft) bedrock, added a few barrel vaults made from what they had dug out, made all the walls 3 foot thick and kept the distance between walls to less than 12ft. The walls themselves are two piles of soft stone held together with lime mortar, with the gap filled up with whatever they could find (including a lot of straw AFAICS.) It's the weight of the infill rubble that keeps the vaults strong.
Now the advantage of having a house that is basically a slightly organised pile of rubble is that, in entropy terms, there just isn't very far for it to degrade. When we bought it there was a huge crack down one wall, and we just filled it up with some more boulders and lime mortar.
All this in a low grade earthqake zone, where the French army tried to burn most of the villages to the ground at one point, but the only effect of this was to make the roofs cave in and make the walls black (as you see when you start drilling holes).
On a related note, the first suspension bridge in the world, in Bristol, UK, is also one of the most stable. Why? Because, like most things that Brunel built, it is overengineered by several orders of magnitude. If he had had a beowulf cluster to do his simulations on, it would be a lot lighter, and would probably have fallen into the Avon Gorge some time ago :-)
Most sites suck
Well, yes, but since my business is letting people look at the sites they want to, whether I think they suck or not, that doesn't really help. I suppose I could try starting the world's first strict xhtml4-only cybercafe, but I suspect that the 3 people in France who would be impressed have Internet connections of their own :-) For the rest of the population, the Internet is a consumer product, and they expect it to work. They don't care how it works, they do care if they can't check their bank balance or whatever.
List of bugs?
I don't have my own, there are various comparisons out there on the net. The last one I found was a real classic, a website where mouseover made the entire paragraph split onto 2 lines. Absolutely nothing in the CSS to justify this. We eventually decided that Mozilla was changing the size of the paragraph element depending on its background colour. The background colour attribute has now been removed, which fixed the problem, so unfortunately I can't demo it, but I showed it to a dozen or so people before they fixed it and everyone agreed that it was quite surreal. I think someone filed it with bugzilla.
Last night I started reading a book by Eric Meyer, and one of the points he makes early on, to my surprise, is that it's a bad move to try to do everything with CSS because browser support is so patchy. And he's a Standards Evangelist for Netscape... so that leaves all of us producing more or less sucky markup depending on how we evaluate the browser mix likely to look at our site this week.
Finally, I don't think the 'MS incompatible to take control of Internet' thing holds water in this particular case. Meyer points out that there have been changes between IE5 and IE6 which move IE closer to W3C standards, and which thus mean that the only way to make a page display properly on both versions of IE is to use the same workaround that was always needed to make pages viewable by IE5 and Mozilla. If that's an example of MS trying to push out other browsers, they need a new strategist :-)
Seriously IE sucks.
Unless you want to do Internet banking, or reliable webchat, or be able to see any site that has only been tested with IE, or which blocks access from any other browser. I use Mozilla in my cybercafe, and there are a significant number of sites that customers want to see that just don't work. I cancelled my own online banking contract because I couldn't get into the site.
And before anyone says that I should change bank, or tell my customers not to use non w3c-compliant websites, or something, can we try to stay in the real world? Most people don't plan their entire life around their choice of browser.
Also, from a webmaster's perspective, I find Mozilla CSS non-compliance anything but easy to work around. Sure, the big things are right, unlike some well-documented areas of IE, but working around a couple of big deviations is a lot easier than working around thousands of small ones. I frequently spend time trying to get a website to render properly with Mozilla, only to find that it works fine with IE, Opera and Konqueror.
All being well, we install IE in the cybercafe next week.
Except that they won't be able to afford to use the place.
How does the microwave tester thing work? Sounds fun...
So just to get our story straight here, AOL sending out lots of unwanted CDs is bad, but the Swiss government sending out lots of CDs is good?
Actually, given the Swiss government's rabid stance on recycling (at one point the genevois were smuggling bags of garbage into France to avoid paying garbage tax), this whole project could be considered to be a form of stealth tax, ie they send you the CD, and if you don't dispose of it properly they fine you...
They aren't the only ones who are insane, although the fact that they also have all that firepower is a little worrying.
I remember going to a particularly paranoid seminar on security, and being told about a group of European missionaries who were exchanging pgp-encrypted emails in some African country, and the secret police asked them why they were doing this, which proved how important it is to encrypt your messages. I thought it proved what a dumb idea it is: anyone screening messages is going to pick out the encrypted ones as potentially interesting, and, even if they can't break the encryption, they can always break your door down, pull out your fingernails or just confiscate your computer.
Sending plain text emails, substituting sensitive words where necessary, strikes me as a far better way to go. Although most of my emails would send even the most conscientious spy to sleep...
It's wierder than that. You are talking about an OS that has been ploughing it's own furrow on custom hardware since before Linux was a twinkle in Linus' eye, so trying to understand it in terms of standard PC hardware is not going to be very useful. As I have explained elsewhere, the whole OS is in the ROM, possibly including the fonts and some applications. So the 'analagous to BIOS' bit of the statement is just not true, IMHO.
The equivalent of the BIOS
Much as I love RISC OS and wish Castle well, this is pretty disingenious. Pretty well the entire OS is in the ROMs: with V3 you could boot the machine into a windowing environment and produce vector graphics with the equivalent of Mac Draw, complete with anti-aliased fonts, and print the result, without even inserting a floppy. I think the applications moved onto disc in v4, but there is still an awful lot more than a BIOS on the ROM.
There was a lot of talk about the industry moving this way a while back. Is it ever going to happen? Sounds eminently fair to me. The result would be far smaller bills for light users, making it easier for people to get started, and that heavy users might think a bit more about how much of this stuff they actually need. If people want to run a worldwide video distribution operation from their bedrooms, that's fine by me, but I don't really see why I should subsidise it...
there will be plenty for everyone
Actually, I doubt it. I think you will find that the company in question has a dozen or so employees. We are not talking Microsoft here. The few people who have ever heard of them are mostly surprised that they are still in business.
Pinching the code was a naughty no-no, but, in terms of a character defence, they were standing up to Microsoft when most people on this list were still happily using DOS, and, for several years, they did a pretty good job, at least in the UK. So I hope that a negotiated settlement can be found.
the only understanding you'll get is from scientific inquiry
Depends exactly what aspect of this accident you are trying to deal with. For reducing the chances of a repeat performance, science is the way to go. For dealing with the existential questions, science is not a lot of help.
Reading this thread from the other side of the Atlantic, I would suggest that there is more going on here than a discussion of the event in reductionist terms.
Yes, 7 people died, and, yes, that's very sad. But how many people died on the DC beltway today? Why are 270 million people going to be grieving for 7 astronauts, rather than for all the other people who died today? Rationally, it makes no sense. But in terms of what the shuttle represents subjectively, expecially to Americans, it makes a lot of sense.
What this accident says to everyone watching is we are all mortal, and analysing the state of the tiles on the Shuttle doesn't begin to bring understanding of that. Death is the new taboo. Death of young people using the latest technology is as bad as it gets in our (shared) culture. That's why people want to pray, or light a candle, or punch someone, or get drunk, or something. Personally, I think praying is one of the more useful options in that list.
Yeah, but wouldn't this have applications in human cloning?
Isn't the idea of cloning yourself that you get to keep the original (ie you) too?
For goodness sake, read the article.
Ooooh, I thought the idea was that one person read the article and we all asked him questions :-)
To all intents and purposes...
I can't say I'm reassured that it would be me rather than a copy of me who stepped out of the teleporter, although the question is probably closer to theology than physics...
How do they know it's the same electron? What distinctive marks are they basing this on? If it's just one electron that disappears and another one that reappears, the applications would be rather more limited. I mean, if I step into a teleportation machine, I would quite like me rather than someone who looks just like me to step out a mile away.