I think there might be a grain of truth in the fact that KDE has very hard time winning the desktop. Gnome has the huge advantage of licensing (LGPL vs. GPL). It doesn't matter how much smoother or better the technology underlying KDE or KDE applications is.
You're right KDE can't win, if you're going to keep shifting the goal posts like that. As long as KDE was not GPL, all you heard was 'Unclean! Unclean!'. Now that KDE is GPL, that isn't good enough either? For heaven's sake make your mind up.
As for preferring KHTML over Mozilla components, I think Apple pointed out last week exactly what a good decision that was. Gecko is a good component, fair enough - but KHTML is at least as good and very much smaller.
The core of the problem is that PCI-SIG has little choice under trademark law -- unless they vigorously persue all infringement, they can potentially lose their trademark. Under trademark law, if the trademark holder knows of infringement and allows it, they loosen their grip on the trademark.
Don't be silly, they didn't have to send him a cease and desist. It would have been easier and cheaper to send him a brief letter thanking him for his work and granting him an explicit license to use the mark, perhaps with some conditions attached.
Don't copy it, but feel free to make similar points. My take on this is this is probably a young, inexperienced lawyer who thought it was big and cool to go in with all guns blazing, and is now (I suspect) desperately trying to dig himself out of the shit. If, on the other hand, PCI-SIG actually instructed him to write in these terms, they deserve everything they get.
Dear Michael Cohen
I appreciate that you're probably getting a lot of grief from strangers today, and probably feel bewildered and a little hurt. You probably feel that you were just doing your job, and that people (including me) are just shooting the messenger.
That's true, of course, to a degree. But in this case it isn't an adequate excuse. Yes, as a lawyer, your job is, in the end, to do what your client instructs. But when your client instructs you to do something extraordinarily foolish and liable to cause grave damage to your clients' own interests, part of a responsible lawyers duty is to councel caution and reflection.
Your clients members are, as a consequence of your action, denied access to a data resource which is vital to them. To replace this resource, which you have by your action denied to them, will cost them many thousands of dollars, delay development of new devices, and cause untold confusion. At the same time, their goodwill and reputation among the technical community on which they depend is in tatters. What possible benefit did you see to your client, and how do you propose that they should go about repairing the damage that has been caused?
After a letter as unnecessarily offensive and aggressive as that which is posted here http://www.yourvote.com/pci/Scanned_.pdf over your signature, saying sorry is not likely to be enough. This isn't a matter of ego, virility, and big swinging dicks. It's a community where people provide resources out of good will and a spirit of co-operation, and you cannot simply go rampaging about in your elephant boots. You (and your clients) have a very great deal of humble pie to eat.
I believe the worlds oldest company would be Stora (From Falun, Sweden), which was founded about 1000 years ago, I also believe you'll find very few people that have reached 1000 years.
Stora dates back to at least 1288. I think there are some companies in the Netherlands which are fourteenth century in date. It would not surprise me if there were companies in the far east which were older.
The Aberdeen Shore Porters Society was founded in 1498 and is still trading - and, indeed, is still in the same business it was in 500 years ago, which says something for consistency. It is reputedly the oldest company in Britain.
Point is, Companies/Corporations don't have a limited lifespan, they can live for nearly forever and keep trying to change legislation for a very very long time, that's one reason I think that businesses should not be considered to be individuals/humans/persons.
While there is room for legitimate discussion of U.S. actions in Guantánamo.
There is no room for legitimate discussion of U.S. actions in Guantánamo. They are absolutely and uncontestably illegitimate, illegal, improper. You cannot conduct a 'war on terror' while acting like that. It robs the United States and its allies of any claim whatever to moral high ground. The US government has demonstrated its total contempt for law, order, judicial procedure, and its obligations under international treaties. That's a very hard place from which to try to criticise other people's bad behaviour.
Because it's garbage-collection based, you can't use D for low-level programming.
Errr, sorry? Why not? I learned my trade on computers which ran LISP right down to the bare metal, with incremental GC. Dammit, the fist WIMP interfaces and the first Ethernet (among other things) were developed on those machines, and, lets face it, a Xerox 1186 in 1986 with 4Mb of core and 80Mb of disk ran complex WIMP programs better and faster than MS Windows does on 2GHz pentium boxes now.
There's nothing wrong with writing low-level stuff in a GC language, although it may be necessary to mark bits of code with a pragma saying don't GC while you're doing this (which is, after all, much what an interrupt routine does on more conventional architectures).
Well, once you have a language that is Turing-complete, it can do anything that any other Turing-complete language can do. Basic theorem of computer science (in the sense as a field of mathematics). And there aren't any languages (that I know of anyway) that are more than that...
.. and why not? Because if it's turing complete it can by definition (give or take the limitations of finite memory) compute all computable functions, and as uncomputable functions are uncomputable, there aren't any programming languages which allow you to compute them. You need to read up on the Entscheidungsproblem.
I disagree. You can put together quite a nice film-based SLR system for around $500-800 or so (camera and lenses -- tripods/bags/filters extra). To get similar quality from a digital SLR would add at least $1000 (probably more) to the price tag. $1000 will buy a lot of film and processing. I am sticking with film for now.
My Canon A1 has sat on the shelf for about two years now; the only time it's been used has been when the digicam (Olympus C2100UZ) was away for repair. Yes, the Canon is a slightly better camera and at the limits takes bettwe pictures - the Olympus is slightly flimsy, its viewfinder isn't really good enough for precise manual focus and its autofocus isn't always trustworthy. But the Olympus is far more versatile and far more useable. I take far more photographs with it. As to the range of photograhic situations it's useful for, I've taken a lot of wildlife photographs, including dragonflies and other insects. I've taken a lot of night-time landscapes, moonlight and starlight shots. I've taken literally hundreds of photographs from and of fast moving boats in bumpy water. And of course I've taken the usual photos of house, friends, pets, etc.
As for resolution, 1600x1200 pixels is good enough for 8"x10" photos and doesn't look too bad blown up even further; obviously it isn't as good as 2000x3000. But for the amateur photographer the digi wins every time. It's lighter and more conenient to carry around, while still having as wide a range of focal lengths (equivalent of 38mm to 400mm) as I've ever carried. It takes snapshots without need for thought; and if you want to set things up to take a proper photograph, control over everything - shutter, aperture, focus, focal length - is there.
You'll get the little Olympus for the same $500ish you were quoting for an SLR kit, but provided you use rechargeable batteries that's all you'll pay. With an SLR, every shot you take costs film and processing, so if like me you take several thousand photographs a year that easily adds up to more money than the camera.
The next camera I buy will have a metal chasis and a proper optical viewfinder. It will also be more optimised for manual focus than the Olympus. But it will definitely be a digital - there's no way I'm going back to film.
IBM badged Tadpole RS6000 laptop, hard disk is dodgy.
I've also got a late model 32k Commodore PET with dual disk drives, but as it isn't British made I don't think of it as part of my collection and will happily swap it for an interesting early British machine.
Yes, I know this is all pretty ggeky. But this is part of our history - in my opinion an important part of our history - and these machines are being thrown into dustbins all the time. Somebody needs to preserve them. So if anyon'e got a Nascom, or an Acorn Model 1 or Acorn Cambridge Workstation that they don't want, let me know.
My favourite error message of all time is a Microsoft Access one, which is quoted in the subject. I can think of nothing more perfect or complete.
As to comments, my best ones always happen while I'm working with Java's date and time stuff, which always seems to me to be pervsely broken.
v = "'" +
new java.sql.Date( value.getTime().getTime()).toString() +
"'";
// Baroque, innit? Why, you say, did // he not use Calendar.getTimeInMillis()? Well, // cos it ain't there, no matter what the friggin docs say.
. ..
catch ( ParseException p)
{
// I really, really would prefer // this didn't happen
. ..
if ( offset == 0)
result.append( "Z");
// zero offset -- excellent, easy.
else
{
// horrible, horrible, oh most horrible.
. ..
/** A simple specialisation of a GregorianCalendar which doesn't make
* a complete balls-up of rendering itself as a string. I have
* commented elsewhere on the wanton overcomplication of the Java
* date handling code... I hope the person who designed it is for
* ever after forced to keep his diary using the output from
* GregorianCalendar.toString() as his date format.
What we do know is that lots of people pirate software, and lot of those people are percieved to be "hackers", and that a lot of "hackers" are percieved to be Linux users. MS knows what is obvious: they won't sell many copies of Office for Linux, that they will still be slammed in the Linux/UNIX world, and that the product will be pirated on a massive scale (or copied in an infringing manner) from the very instant it is available.
I would be careful saying that too often. There's no evidence that Linux users are more likely to pirate software than windoes users, quite the contrary. There is no pirate softwarea at all on any of my forty-odd machines. I think you'll find the same is true of many, probably most, other people who use Linux exclusively. There's plenty of paid for software on my machines - I have paid-for, boxed copies on the shelf of twelve different Linux distrributions, five different Linux games, three different Linux office productivty suites, one IDE, and three DBMSs.
The suggestion that Linux users are likely to be dishonest is certainly untrue and libelous, and sooner or later you are going to meet someone who takes serious offence at it.
Is Microsoft actually dumb enough to write memo after memo about something they now have admitted is their biggest threat and allow all of these memos to leak so the opposition can read them?
It isn't dumb to analyse your competition, nor to plan responses to news about your competitors successes. Microsoft's problem is that it is a very large, and very distributed organisation, and at least some of its employees are disaffected. So it's inevitable that any widely distributed memo will leak. Microsoft, obviously, know this too by now, so it has to be assumed that this memo was written in the expectation that it would leak.
I was never sure about the first Halloween memo. The more that are "discovered" the more I wonder if these are truly from M$ (they must be released by our old friend, Mr. Source, or Reliable to those that know him well).
Microsoft acknowledged that the first Halloween Memo was genuine.
This latest memo says nothing more than you would expect any sensible company to be saying at this time. ESR is right, of course, to point out that it shows Microsoft is on the defensive. You would expect a well run company to be, at this moment when so many major and influential customers are publicly looking at the competition. I agree with other posters that on this occasion ESR's annotations look shrill and are not, in my opinion, likely to sway unbiased readers in our favour.
Uhhhh. What is "PERL"? Closest match I found is "Perl", a scripting language heavily used in system maintenance procedures. Kind of important if you want to, say, install packages and stuff. (Actually I don't know about RH8, which is what we're presumably talking about here, but a lot of Debian packages require a minimal Perl install to support the packaging scripts.)
Oh - for future reference, saying "PERL" rather than "Perl" is a lot more damaging to your credibility than any mere spelling mistake I can think of. (:
PERL is an acronym, standing (misleadingly) for 'Practical Extraction and Reporting Language'. Larry Wall subsequently decided he didn't want it to be an acronym any more and changed history. Thus anyone who writes 'PERL' is old enough to remember PERL 1 or PERL 2, while anyone who writes 'Perl' is either a revisionist or a clueless newbie.
Oh, and anyone who things it's a good idea to hack bastardised forms of awk(an acronym, named after its developers Aho, Kernighan and Weinberger) and sed(a contraction of stream editor) into one process doesn't understand UN*X.
In the UK where similar ideas are being considered the issue is to be able to charge more for roads that are more congested, to reduce the risk of gridlock in cities and at chokepoints on the highway network by providing people with incentives to use less busy roads or other forms of transport for urban travel.
Essentially, minor rural roads would be free and rural roads generally would be very low cost whereas innner city roads would be extremely highly taxed (in US prices, up to about seventy cents a mile).
The idea isn't nearly as stupid as it first sounds.
Because being a policeman myself, I know that by the time a search warrant is signed off by a judge and executed (around a week), the trash will be long gone. So, the policeman have a perfectly valid arguement.
The councilman have every right to call foul play, because the police are an investigation bureaucracy devoted to helping people (legally), while the reporters are going through garbage in order to report what bills the councilman paid last week (illegally).
Being a policeman yourself, you'll know that a policeman without a warrant is just a citizen like any other, and if it's good and leagal for you it's good and legal for anyone else.
The problem is that most professors want to see your *work*. If you just gave a few steps and -voila- an answer, they usually don't appreciate it.
This is true. I know, I used to drive my teachers mad...
I'm extremely dyslexic, and have extreme difficulty writing. So in maths classes I just used to write down the answers with no working. If I tried to write down the working, I couldn't keep up with the class. This was in the days before calculators, so it wasn't a case of doing things on the calculator, I just did them in my head. Most of the time I got the answers right - I was extremely good at maths - but my teachers hated it.
The entire gameplay is on that one screen, moving toward the woman while dodging arrows. The arrows fall in a random, unpredictable, unlearnable pattern.
OK, OK, OK. . .
Can anyone think of anything more seriously uncool than admitting you've played that game long enough to work that out?
I have some (very few) visual memories from very early. I can remember being carried into my parents bedroom. This is a purely visual image and comes from a house we left when I was 18 months old. I can remember a bottle of orange juice on my mother's dressing table. That could be any time up until I was about four, but given that my viewing angle is upwards I think it's from when I was two. I have a number of visual memories specific to a house we left when I was four. These are all memories from my viewpoint of things we don't have photographs of, so I'm confident they are real memories not things I've been told about. These are all sort of like snapshots - I can see the scene visually and I can describe it, and in some of them I know what's going on, but there's no 'story', no connected sequence of memories.
From the next couple of years I've got some sequential memories which I'm really sure are genuine personal memories (i.e. they're not things I've been told about and no-one else would have remembered them to be able to tell me about them).
I've discussed early memory with a number of people and I think it's very variable. Some people remember stuff from much earlier than others.
Herein lies the problem: an interface that requires relearning can, at least in a sense, be said to be flawed.
I won't deny there's a level of truth in what you say. For example, we're (almost) all still using qwerty keyboards not because they're a good keyboard design, and not, on the whole, because we can use them very efficiently, but because it's what people have used since the beginning of typewriting and the retraining cost of change is just too high.
But hey, wait a minute, the cost of relearning a user interface style is not necessarily high. Pretty much every computer game you play uses a non-standard user interaction style, and many of these games have really quite radically different user interaction styles. Many of these user interaction styles are far more in tune with current psychological and HCI research than the tired variants of WIMP noow used by Mac/Windows/KDE/Gnome et cetera. If users could not quickly learn these styles the games would go unsold and unplayed.
So I don't believe that a radically new user interaction style is impossible, provided that it is highly intuitive, highly consistent and easy to learn. The problem will be (as it always is when changing a user interface) that initially there won't be enough apps ported so that you will have to mix between older apps with an older interaction style and the new apps, robbing the system - at least initially - of the consistency and predictability which must ultimately be an important feature.
One possible strategy would be to take a widely used toolkit such as KDE and write your new user interface library to have the same API. The problem with this is that you may end up compromising your user interface design in order to support some features offered by the 'old' toolkit which don't really fit your new model.
About four years ago I was stupid enough to lose control of a classic sportscar on a mountain road and I broke my back. This, as you can imagine, doesn't feel too good. A very competent ambulance crew scraped me up and took me down the mountain. And there I was, lying with a broken back in the back of an ambulance driving quite fast down a twisting, bumpy mountain mountain road thinking that I was probably going to be paralysed for the rest of my life...
And he's right, you know. That very competent ambulence crew shot me full of morphine (which is basically the same stuff as heroin) and I felt absolutely wonderful. Absolutely wonderful. I'm not the least surprised that people get addicted to it.
Oh, yes - thanks at least partly to that very competent ambulance crew and great medical staff at Ayr hospital, I made a full recovery.
Are you on crack? Would you *want* to have to handle 200 systems if they all went down at once?
If all your servers go down at once you haven't done your job properly. Where's your UPS? Where's your backup generator? Where's your disaster recovery plan? So far the amount of unscheduled downtime on my servers in 2002 is 0 seconds. What's yours?
Slashdot Mirror
You're right KDE can't win, if you're going to keep shifting the goal posts like that. As long as KDE was not GPL, all you heard was 'Unclean! Unclean!' . Now that KDE is GPL, that isn't good enough either? For heaven's sake make your mind up.
As for preferring KHTML over Mozilla components, I think Apple pointed out last week exactly what a good decision that was. Gecko is a good component, fair enough - but KHTML is at least as good and very much smaller.
Don't be silly, they didn't have to send him a cease and desist. It would have been easier and cheaper to send him a brief letter thanking him for his work and granting him an explicit license to use the mark, perhaps with some conditions attached.
Don't copy it, but feel free to make similar points. My take on this is this is probably a young, inexperienced lawyer who thought it was big and cool to go in with all guns blazing, and is now (I suspect) desperately trying to dig himself out of the shit. If, on the other hand, PCI-SIG actually instructed him to write in these terms, they deserve everything they get.
Dear Michael Cohen
I appreciate that you're probably getting a lot of grief from strangers today, and probably feel bewildered and a little hurt. You probably feel that you were just doing your job, and that people (including me) are just shooting the messenger.
That's true, of course, to a degree. But in this case it isn't an adequate excuse. Yes, as a lawyer, your job is, in the end, to do what your client instructs. But when your client instructs you to do something extraordinarily foolish and liable to cause grave damage to your clients' own interests, part of a responsible lawyers duty is to councel caution and reflection.
Your clients members are, as a consequence of your action, denied access to a data resource which is vital to them. To replace this resource, which you have by your action denied to them, will cost them many thousands of dollars, delay development of new devices, and cause untold confusion. At the same time, their goodwill and reputation among the technical community on which they depend is in tatters. What possible benefit did you see to your client, and how do you propose that they should go about repairing the damage that has been caused?
After a letter as unnecessarily offensive and aggressive as that which is posted here http://www.yourvote.com/pci/Scanned_.pdf over your signature, saying sorry is not likely to be enough. This isn't a matter of ego, virility, and big swinging dicks. It's a community where people provide resources out of good will and a spirit of co-operation, and you cannot simply go rampaging about in your elephant boots. You (and your clients) have a very great deal of humble pie to eat.
Yours Sincerely
Simon Brooke
Chief Technical Officer, Scaffie Ltd.
The Aberdeen Shore Porters Society was founded in 1498 and is still trading - and, indeed, is still in the same business it was in 500 years ago, which says something for consistency. It is reputedly the oldest company in Britain.
There are four companies in the United States, even, which date back to before independence.
Absolutely agreed.
There is no room for legitimate discussion of U.S. actions in Guantánamo. They are absolutely and uncontestably illegitimate, illegal, improper. You cannot conduct a 'war on terror' while acting like that. It robs the United States and its allies of any claim whatever to moral high ground. The US government has demonstrated its total contempt for law, order, judicial procedure, and its obligations under international treaties. That's a very hard place from which to try to criticise other people's bad behaviour.
Errr, sorry? Why not? I learned my trade on computers which ran LISP right down to the bare metal, with incremental GC. Dammit, the fist WIMP interfaces and the first Ethernet (among other things) were developed on those machines, and, lets face it, a Xerox 1186 in 1986 with 4Mb of core and 80Mb of disk ran complex WIMP programs better and faster than MS Windows does on 2GHz pentium boxes now.
There's nothing wrong with writing low-level stuff in a GC language, although it may be necessary to mark bits of code with a pragma saying don't GC while you're doing this (which is, after all, much what an interrupt routine does on more conventional architectures).
.. and why not? Because if it's turing complete it can by definition (give or take the limitations of finite memory) compute all computable functions, and as uncomputable functions are uncomputable, there aren't any programming languages which allow you to compute them. You need to read up on the Entscheidungsproblem.
Not so. Sorry to be boring.
Before C came B, before B came BCPL, before BCPL came CPL (except it was never finished), and (arguably) before CPL came Algol.
My Canon A1 has sat on the shelf for about two years now; the only time it's been used has been when the digicam (Olympus C2100UZ) was away for repair. Yes, the Canon is a slightly better camera and at the limits takes bettwe pictures - the Olympus is slightly flimsy, its viewfinder isn't really good enough for precise manual focus and its autofocus isn't always trustworthy. But the Olympus is far more versatile and far more useable. I take far more photographs with it. As to the range of photograhic situations it's useful for, I've taken a lot of wildlife photographs, including dragonflies and other insects. I've taken a lot of night-time landscapes, moonlight and starlight shots. I've taken literally hundreds of photographs from and of fast moving boats in bumpy water. And of course I've taken the usual photos of house, friends, pets, etc.
As for resolution, 1600x1200 pixels is good enough for 8"x10" photos and doesn't look too bad blown up even further; obviously it isn't as good as 2000x3000. But for the amateur photographer the digi wins every time. It's lighter and more conenient to carry around, while still having as wide a range of focal lengths (equivalent of 38mm to 400mm) as I've ever carried. It takes snapshots without need for thought; and if you want to set things up to take a proper photograph, control over everything - shutter, aperture, focus, focal length - is there.
You'll get the little Olympus for the same $500ish you were quoting for an SLR kit, but provided you use rechargeable batteries that's all you'll pay. With an SLR, every shot you take costs film and processing, so if like me you take several thousand photographs a year that easily adds up to more money than the camera.
The next camera I buy will have a metal chasis and a proper optical viewfinder. It will also be more optimised for manual focus than the Olympus. But it will definitely be a digital - there's no way I'm going back to film.
If no-one is hiring, hire yourself. The world doesn't owe you a living.
No, I haven't got and never had a Spectrum. I'm looking for an original 16k rubber key model, but seeing they were so common it isn't a priority.
I've also got a late model 32k Commodore PET with dual disk drives, but as it isn't British made I don't think of it as part of my collection and will happily swap it for an interesting early British machine.
Yes, I know this is all pretty ggeky. But this is part of our history - in my opinion an important part of our history - and these machines are being thrown into dustbins all the time. Somebody needs to preserve them. So if anyon'e got a Nascom, or an Acorn Model 1 or Acorn Cambridge Workstation that they don't want, let me know.
As to comments, my best ones always happen while I'm working with Java's date and time stuff, which always seems to me to be pervsely broken.
v = "'" + new java.sql.Date( value.getTime().getTime()).toString() + "'";
. . .
catch ( ParseException p)
{
. . .
if ( offset == 0)
else. . .
* a complete balls-up of rendering itself as a string. I have
* commented elsewhere on the wanton overcomplication of the Java
* date handling code... I hope the person who designed it is for
* ever after forced to keep his diary using the output from
* GregorianCalendar.toString() as his date format.
I would be careful saying that too often. There's no evidence that Linux users are more likely to pirate software than windoes users, quite the contrary. There is no pirate softwarea at all on any of my forty-odd machines. I think you'll find the same is true of many, probably most, other people who use Linux exclusively. There's plenty of paid for software on my machines - I have paid-for, boxed copies on the shelf of twelve different Linux distrributions, five different Linux games, three different Linux office productivty suites, one IDE, and three DBMSs.
The suggestion that Linux users are likely to be dishonest is certainly untrue and libelous, and sooner or later you are going to meet someone who takes serious offence at it.
It isn't dumb to analyse your competition, nor to plan responses to news about your competitors successes. Microsoft's problem is that it is a very large, and very distributed organisation, and at least some of its employees are disaffected. So it's inevitable that any widely distributed memo will leak. Microsoft, obviously, know this too by now, so it has to be assumed that this memo was written in the expectation that it would leak.
Microsoft acknowledged that the first Halloween Memo was genuine.
This latest memo says nothing more than you would expect any sensible company to be saying at this time. ESR is right, of course, to point out that it shows Microsoft is on the defensive. You would expect a well run company to be, at this moment when so many major and influential customers are publicly looking at the competition. I agree with other posters that on this occasion ESR's annotations look shrill and are not, in my opinion, likely to sway unbiased readers in our favour.
PERL is an acronym, standing (misleadingly) for 'Practical Extraction and Reporting Language'. Larry Wall subsequently decided he didn't want it to be an acronym any more and changed history. Thus anyone who writes 'PERL' is old enough to remember PERL 1 or PERL 2, while anyone who writes 'Perl' is either a revisionist or a clueless newbie.
Oh, and anyone who things it's a good idea to hack bastardised forms of awk (an acronym, named after its developers Aho, Kernighan and Weinberger) and sed (a contraction of stream editor) into one process doesn't understand UN*X.
Essentially, minor rural roads would be free and rural roads generally would be very low cost whereas innner city roads would be extremely highly taxed (in US prices, up to about seventy cents a mile).
The idea isn't nearly as stupid as it first sounds.
Being a policeman yourself, you'll know that a policeman without a warrant is just a citizen like any other, and if it's good and leagal for you it's good and legal for anyone else.
No, it isn't. Not when you're posting to Slashdot, it isn't. If you think it is, it's very easy to understand why you think you haven't been hacked.
This is true. I know, I used to drive my teachers mad...
I'm extremely dyslexic, and have extreme difficulty writing. So in maths classes I just used to write down the answers with no working. If I tried to write down the working, I couldn't keep up with the class. This was in the days before calculators, so it wasn't a case of doing things on the calculator, I just did them in my head. Most of the time I got the answers right - I was extremely good at maths - but my teachers hated it.
OK, OK, OK. . .
Can anyone think of anything more seriously uncool than admitting you've played that game long enough to work that out?
From the next couple of years I've got some sequential memories which I'm really sure are genuine personal memories (i.e. they're not things I've been told about and no-one else would have remembered them to be able to tell me about them).
I've discussed early memory with a number of people and I think it's very variable. Some people remember stuff from much earlier than others.
I won't deny there's a level of truth in what you say. For example, we're (almost) all still using qwerty keyboards not because they're a good keyboard design, and not, on the whole, because we can use them very efficiently, but because it's what people have used since the beginning of typewriting and the retraining cost of change is just too high.
But hey, wait a minute, the cost of relearning a user interface style is not necessarily high. Pretty much every computer game you play uses a non-standard user interaction style, and many of these games have really quite radically different user interaction styles. Many of these user interaction styles are far more in tune with current psychological and HCI research than the tired variants of WIMP noow used by Mac/Windows/KDE/Gnome et cetera. If users could not quickly learn these styles the games would go unsold and unplayed.
So I don't believe that a radically new user interaction style is impossible, provided that it is highly intuitive, highly consistent and easy to learn. The problem will be (as it always is when changing a user interface) that initially there won't be enough apps ported so that you will have to mix between older apps with an older interaction style and the new apps, robbing the system - at least initially - of the consistency and predictability which must ultimately be an important feature.
One possible strategy would be to take a widely used toolkit such as KDE and write your new user interface library to have the same API. The problem with this is that you may end up compromising your user interface design in order to support some features offered by the 'old' toolkit which don't really fit your new model.
About four years ago I was stupid enough to lose control of a classic sportscar on a mountain road and I broke my back. This, as you can imagine, doesn't feel too good. A very competent ambulance crew scraped me up and took me down the mountain. And there I was, lying with a broken back in the back of an ambulance driving quite fast down a twisting, bumpy mountain mountain road thinking that I was probably going to be paralysed for the rest of my life...
And he's right, you know. That very competent ambulence crew shot me full of morphine (which is basically the same stuff as heroin) and I felt absolutely wonderful. Absolutely wonderful. I'm not the least surprised that people get addicted to it.
Oh, yes - thanks at least partly to that very competent ambulance crew and great medical staff at Ayr hospital, I made a full recovery.
If all your servers go down at once you haven't done your job properly. Where's your UPS? Where's your backup generator? Where's your disaster recovery plan? So far the amount of unscheduled downtime on my servers in 2002 is 0 seconds. What's yours?