If we are going to argue semantics, a better title would be geisha or coutesan. It's a pity these things only come out when a company very publicly crashes and burns.
This sort of requirement has settled down to MS Word now, but not long ago technical staff that could work out how to use any word processing package in detail with less than five minutes with a manual (or ten without) were not considered unless they listed a particular word processing package on their resume. I had about twelve listed on mine for such situations, from Chiwriter up. All this is irrelevant, however, when you submit the resume as a PDF file and the employement agent doesn't know how to read it.
Which is a pity, we need less managers who focus on their "power animal" and whatever that latest management fad is - while giving away half the companies assets to their mates. It doesn't matter how good a programmer you are, there are secretaries that get paid more than you do. They get to stay in five star accomodation during the week, get flown home on weekend, and for a work expense they could claim little frilly... on second thoughts they can keep their jobs - but shareholders should know that they are paying for that sort of thing.
A little less corruption and more competance in business would be a good thing. I now work for a company with very transpanent accounting, and have a compentent boss. A previous boss failed to supply electricity to a major city for over a month, but I'm sure "Quality" was maintained.
It won't be diamond at those temperatures
on
The Diamond Age
·
· Score: 1
you have this chip that will now run at 2k instead of 200 degrees
Long before you get to that point you'll have nothing but graphite. Long before you have graphite the semiconducting properties will be gone and you will have an insulator. If you look at Silicon, it melts at well over 1000 degrees celcius, but before you get to 100 celcius your doped silicon is going to let so little current through that it may as well be an insulator.
Semiconductor behaviour is very much temperature dependant.
Also, diamond is what is know as a "metastable" material. Carbon really wants to be graphite at atmospheric pressure, and given enough temperature and time it will be. At room temperature it will take enormous amounts of time to transform to graphite, but as the temperature increases the time taken to transform drops to hours, minutes or seconds. At very high pressures carbon wants to be diamond - which is why it is so difficult to make the stuff.
Material Scientists used to laugh at a very lame James Bond plot - smuggling diamonds inside corpses and getting the diamonds out of the cremated ashes. If anyone actually tried that they would just get very expensive bits of graphite.
Re:degree=a point in any scale
on
The Diamond Age
·
· Score: 1
Thus, using the word 'degrees,' or the term 'degrees Kelvin,' may be accurate, but it is not as accurate or concise as saying 'Kelvins.'
It must be an engineer VS chemist thing, or geographically altered english, but I've heard them called both, and everone has know what has been talked about. The symbol is of course just "K" with no degree symbol.
Diamond doesn't melt - it transforms to graphite
on
The Diamond Age
·
· Score: 1
diamonds melt at ~4000 K
Diamond doesn't melt - it transforms to graphite at below 500 celcius at atmospheric pressure in a very short period of time. Anyone that wants an exact figure should look up the phase diagram - I don't have one in the house. There are a lot of materials that change crystal structure before they melt.
Artificial diamond of up to 7mm in diameter has been available commercially for a few years now - but bigger crystal sizes extend the range of uses - like the semiconductors in the article.
Imagine a processor that will run at many times the current CPU upper temps and not blink
Diamond can't get as hot as you would think before it transforms into graphite - there's no way a diamond CPU can run at double the temp of silicon in degrees kelvin, or three times in degrees celcius.
Each time I used info I had to type "info info" and try to work out how to get past the first page because the writers were originally too elitist to write a man page.
pinfo is usable and behaves like a web browser (specificly lynx) - hit an arrow key on a link in pinfo and you go there, no odd commmands you forget between uses.
One thing I dislike is project like grub - where they have a lot of text on thier web page saying how bad their competitor is and why the manual is in info format, but the links to documentation are all broken.
# The fact that 'weird places' means that there are a half-dozen places for binaries to go (/bin,/sbin,/usr/bin,/usr/sbin/, etc...)
#... in fact, I find the whole/usr heirarchy annoying. Why was that necessary? Weren't the six other folders for binaries enough?
# Don't even bring up/opt!
#...or/usr/share!
I see the problem here - it's a unix clone.
There are a lot of good reasons to put binaries in different places on mulituser (/usr/bin,/bin ,/sbin) and networked (/usr/share ,/usr/local) systems with specific large appications that justified buying the system in the first place (/opt). Linux tries to be like those, unstead of just throwing the binaries wherever the user wants to put them (MSDOS) or in one single directory (C:\Program Files\)./root is not under home so that you can always get to it when even when/home is on another machine down the corridoor that is not accessable right now.
I suppose I see this because I look after machines running linux, solaris and AIX - with csh, ksh and bash. If something is in sh it will just run. The expensive software that runs on these machines is cross platform - and is written to use sh and other things that it can expect to find in the same spot on all those platforms.
Too much dependance on editing configuration files by hand. While this can and should always be an option, I've had to do it too many times where it was obvious that the feature should have been accessible through a gui.
When it really, really has to be accessable in single user mode you do not want to only know the GUI way. I started with GUI's early (atari ST) and couldn't get anything serious done until I bought a program that gave me a decent command line on that platform. Sometimes it's better to keep it simple - and there's not much simpler than a list of options in a text file.
Inconsistent location of files./usr ,/usr/local ,/bin ,/sbin, and the like are not intuitive and not consistently used either
annoys me too - it comes down to ignorance by those that put things where they put them. I have sixteen machines that someone set up to share/usr/local/ , while each machine has local stuff on/usr/share/ . Installing new stuff is a pain - it always wants to be in/usr/local/ and sometimes keeps machine specific things there on the assumption that everone would know that local is supposed to be local.
At high school I was taught calculus BY ROTE by a lazy teacher (who would leave five minutes into each lesson) and bad textbooks - but I have a good memory so I easily made it into university without having a clue about calculus. Books like "Advanced Engineering Mathematics" by Kreysig, and another by Thomas & Finney taught me what calculus was from scratch (and suddenly the basic concepts seemed so incredibly simple), by giving some real world examples. There are a lot of similar books designed to teach engineers how motion is described by mathematics, so just pick a big second hand book store and grab any with lots of examples and diagrams printed within the last fifty years.
All the texts I've seen aimed at school students have no relation to the physical world at all - mainly I suspect because many high school teachers don't know how the mathematics they teach can have any possible application. If an engineer or a mathematition chooses a text for a course they are less likely to pick something that is there just to tick a box labelled "student knows calculus", and are more likely to pick something that leads to understanding.
Then again - I was a lazy student that never bothered to learn my times table - I couldn't see any point in learning more than the prime numbers.
Combine that with all his historic enemies due to Iran/Contra
Considering that those enemies would be the US taxpayer and all of the military allies of the USA, it is a wonder he every was put in a position of reponsibility ever again.
Then again, I also think letting someone that came from the corrupt Nixon government run the largest defence force in the world is a bit of a bad call - no matter how good a wrestler he used to be.
Explorer dying is generally worse than the GNOME panel crashing
Who cares if the gnome panel crashes, you still have the window manager so you can just click the mouse on the desktop to bring up some apps (like an xterm to run things from).
tying functionality into Explorer instead of doing it the right way, in the kernel
Putting more things into the kernel is not a good thing - things that can be run in unpriveledged userspace are best done there. How many exploits do you want to see today?
MS is steadily progressing towards a solid and dependable OS, with features aproaching those of CP/M (a low-end multiuser OS that predates DOS), but with the advantage of greater usability and the ability to run on better hardware than CP/M ever did. Win2003 is now a fully multi-user operating system, unlike 2000 or NT4. I believe that new versions of Exchange actually have a feature that allows you to change the postmaster password - so that companies no longer have to trust their former sysadmins for life (I pray that places I've worked never get hacked - since they can't change that password that will make everyone that has known it a suspect).
As for stability, anyone that has worked with a mixed bunch of servers knows the answer to the MS claims about that. In contrast I've had to reboot a solaris box that is under heavy load twice in the last four months due to hassles with external SCSI hardware - that is the sort of reliability a server OS has to have. Linux can do that too if set up correctly. The uptimes on the 26 linux servers and cluster nodes I look after date back to the last prolonged power failure, or hardware upgrades on some boxes. With windows boxes I've found it better if you reboot them once a week whether they seem to need it or not - solves most of those memory leak problems.
If they think they can get around the cost of decent sfx with CGI, I fear they're mistaken.
Babylon5 was cheap as SF TV goes due to all the CGI. Red Dwarf was even cheaper - and Space Island One must have had a tiny budget but still worked well.
If you consider Canadian SF - The cube didn't have a big budget.
Blakes 7 with good sfx would seem wrong somehow. If were willing to suspend disbeleif enough for FTL travel we may as well suspend disbelief that a red platic esky is a high tech tool kit. The strength of the show was in the characters (well those that were at least 2D), and since it worked well as a radio play recently, lots of expensive visual effects are not required.
I think it stands as one of the few self-consistant SF TV programmes ever made. I also liked the digs at the Trek utopian federation - the trek symbol at 90 degrees and the federation as a police state.
I wonder what a material scientist would reply to that
Well, I used to be one, but later study flung me into the realms of numerical computing.
The author of the article has made a common error - confusing strength to weight ratio with strength.
Stength is the amount of force a given cross-section of material can take before it fails. It's given in units of Newtons per millimeter squared (which is a megapascal) or pounds per squre inch. Failure is either when the material is bend permanantly out of shape (yeild strength) or broken (ultimate tensile strength - if it is broken by pulling).
Now a piece of the cheapest steel wire you can can get is going to be stronger than the same cross section of the best bamboo, spider silk or carbon fibre re-inforced plastic you can get. With those light materials you just make it thicker if you want to carry the same load, and you still end up with something ligher than the smaller piece of steel. Those materials have a better strength to weight ratio than steel, but you have thinker peices of material taking the load.
It all comes down to the purpose - you don't want planes or bikes to weigh a lot so you only make the bits that have to be really strong out of steel - say axles and spokes.
longtitudal force
That's along the length of the bamboo stems, the direction the fibres run - the bamboo will be strong in that direction under both tension and compression. Under compression it will buckle if you push it hard enough, because it isn't very stiff. The answer is just to use more bamboo.
material under increasing force goes through ellastic transformation
Polymers (eg. cellulose in bamboo) behave differently to metals - their behaviour depends on time as well as load, and something that looks like it has permanantly bent out of shape can sometimes slowly return to its old shape. The change from fairly elastic behaviour (double the force and get double the stretch) to plastic behaviour is hard to spot. Metals behave elasticly, you know
I would think that the joints are the weak point
Yes, you would just have to design the bike based on the strength of the joint. The same applies with metal bikes, you have to design based on the strength of the weld joint.
If the two rods were made of bamboo, they would be straight and thus more prone to bending to the outside once the brakes are pressed.
That means that the bamboo bike would have to be designed a different way to a metal bike. You have to design to use the material, as well as choose the material for the design.
Re:Before all the flamers get in.
on
Qt On DirectFB
·
· Score: 1
This means I can't develop any non-free apps at all, since QT is GPL
You can, you just need to buy a licence - just like you do when you develop with any commercial software.
Your licence money will not only help push Qt forward, it will also, quite literally, feed the trolls.
This is why my users don't use XFree86 on win*
on
Qt On DirectFB
·
· Score: 1
I don't want a whole bloody desktop with my remote X, thanks. I want a window.
That is exactly why my users keep using an old and crappy version of PC-Xware that can only do 256 colours instead of using XFree86. When there was a problem with an app displaying on the screen they were happy to live with instability instead of running the X server in full screen mode via xdm. They want their windows, and they want them managed the same way their local windows are managed.
I really like X - I prefer to even use twm to windows, but the furthur accross platforms apps stretch the better. My job would be very difficult without *nix apps ported to windows (eg. putty (ssh) and vi), and the more apps that can run on embedded systems the better. Having a web server on a photocopier is a very useful thing, and as more apps are ported things will only get better.
What they are doing is illegal in my country - in exactly the same way the scams where companies send you out invoices for domain name renewal for domains that they don't own is illegal.
Two charges would apply as far as I know - fraud and demanding money with menaces.
improving the X architecture to provide things like AA fonts, better image caching, faster (and more useful) graphics operations and features (like alpha channel)
Some of those things are what Raster of Enlightenment fame has been working on for some time (evas, imlib2, fnlib etc.). He's said a few interesting things about possibilities and limitations of X over the past five years, many of which can still be read.
I also think that Linux badly needs a unified display architecture with a common set of libraries
It would be nice, svgalib has been left behind for years, full screen X replaces it in a lot of situations.
For me, I'll just be happy with a working Xnest or other solution that lets me do overlays of different bit depths. Must read more about X, and look at code.
Yet strangely the many layers between X apps and the display architecture manages to slow down the GUI. (app, widget libraries,.., window managers, session managers..
Those things listed above are not part of X itself, but are seperate applications. If gnome is slowing everthing down, it isn't the fault of X. If an application itself isn't responsive, it isn't the fault of X. If a widget library loads thai laguage support before it will let anything happen with latin fonts, (or even if it isn't going to display any text at all) it isn't the fault of X, it is a badly designed widget set. If the window manager is slowing everything down, use sensible settings (turn those 32 different background images into one if you need the speed) or use a different theme or a different window manager, there are dozens, from twm up. Most of the cruft mentioned above can just be turned off or discarded. Running OpenGL animations like atlantis as the root background window at high resolution will noticeably slow down just about anything, and similarly other bits of less obvious eye candy will also do so. To get a similar response to a windows machine that isn't really doing anti-aliased fonts, turn off anti-aliased fonts.
As an example of where X is useful on a low end PC: On Friday on the local machine I was running concurrently 3 copies of X (one nested) with three different window managers with apps from 8 different machines with three very different operating systems. The machine was responsive enough that I don't notice any lag at all, but I wouldn't want to use a similar machine for 3D games newer than Quake][. On a usual day I would run no more than 2 copies of X, and two window managers, and when a very old, very expensive but very useful scientific application is upgraded I will only need one. This application will not run on the local machine at better than glacial speed, let alone disk size considerations, and having a top end PC for everyone that needs to use that application would cost a lot, but no more than a few percent of the licencing costs for each new copy that would be needed.
My co-workers who use windows also use a version of X to get their work done, only they can't work in both 8 bit and 32 bit without closing applications and going into the windows control panel. If their windows video driver and hardware supported overlays they wouldn't need to do that. The desktops run the range from pentium 166 with 32MB up to dual processor atholons, and all run at a good pace in X, since the real work is done on a server and the local machine just does the display for the serious applications. Local stuff, of course, isn't done at the same speed, and the local stuff decides the OS (eg. MS if you need powerpoint). XFree86 on windows in not yet at a stage where it would be as useful to them as the commercial X servers, but it is rapidly getting there. XFree86 runs on five of the machines they use every day, and happily feeds them the windows they need.
To sum up - X and MS Windows are very different things, about a third of the people I work with use both at once on the same screen.
True - it changes resolution, hence the word resolution used in the subject heading. As for changing colour depth, you do the same thing that is done in windows, switch off the visual and start it up again with a differnet colour depth. The big difference with MS is that it doesn't call all your apps and it remebers where to put the windows.
You can theretically run a lower resolution in X under Xnest (Xnest -ac -depth 8:1) but need the right hardware and drivers. Has anyone had any luck using Xnest this way? On what sort of hardware did it work?
Its no where near windows 95 even in 2003.
It's very, very different. I myself prefer to keep the real screen size as a virtual desktop when I zoom in, so perhaps the developers didn't see what you are asking for as a worthwhile feature. An application (eg. any Loki game) can resize the screen in X quite happily, but the user is limited to stopping and restarting the server with different options.
Slashdot Mirror
This sort of requirement has settled down to MS Word now, but not long ago technical staff that could work out how to use any word processing package in detail with less than five minutes with a manual (or ten without) were not considered unless they listed a particular word processing package on their resume. I had about twelve listed on mine for such situations, from Chiwriter up. All this is irrelevant, however, when you submit the resume as a PDF file and the employement agent doesn't know how to read it.
A little less corruption and more competance in business would be a good thing. I now work for a company with very transpanent accounting, and have a compentent boss. A previous boss failed to supply electricity to a major city for over a month, but I'm sure "Quality" was maintained.
Semiconductor behaviour is very much temperature dependant.
Also, diamond is what is know as a "metastable" material. Carbon really wants to be graphite at atmospheric pressure, and given enough temperature and time it will be. At room temperature it will take enormous amounts of time to transform to graphite, but as the temperature increases the time taken to transform drops to hours, minutes or seconds. At very high pressures carbon wants to be diamond - which is why it is so difficult to make the stuff.
Material Scientists used to laugh at a very lame James Bond plot - smuggling diamonds inside corpses and getting the diamonds out of the cremated ashes. If anyone actually tried that they would just get very expensive bits of graphite.
Artificial diamond of up to 7mm in diameter has been available commercially for a few years now - but bigger crystal sizes extend the range of uses - like the semiconductors in the article.
Good stuff, but heat is very bad for diamond.
pinfo is usable and behaves like a web browser (specificly lynx) - hit an arrow key on a link in pinfo and you go there, no odd commmands you forget between uses.
One thing I dislike is project like grub - where they have a lot of text on thier web page saying how bad their competitor is and why the manual is in info format, but the links to documentation are all broken.
There are a lot of good reasons to put binaries in different places on mulituser (/usr/bin, /bin , /sbin) and networked (/usr/share , /usr/local) systems with specific large appications that justified buying the system in the first place (/opt). Linux tries to be like those, unstead of just throwing the binaries wherever the user wants to put them (MSDOS) or in one single directory (C:\Program Files\). /root is not under home so that you can always get to it when even when /home is on another machine down the corridoor that is not accessable right now.
I suppose I see this because I look after machines running linux, solaris and AIX - with csh, ksh and bash. If something is in sh it will just run. The expensive software that runs on these machines is cross platform - and is written to use sh and other things that it can expect to find in the same spot on all those platforms.
All the texts I've seen aimed at school students have no relation to the physical world at all - mainly I suspect because many high school teachers don't know how the mathematics they teach can have any possible application. If an engineer or a mathematition chooses a text for a course they are less likely to pick something that is there just to tick a box labelled "student knows calculus", and are more likely to pick something that leads to understanding.
Then again - I was a lazy student that never bothered to learn my times table - I couldn't see any point in learning more than the prime numbers.
Then again, I also think letting someone that came from the corrupt Nixon government run the largest defence force in the world is a bit of a bad call - no matter how good a wrestler he used to be.
MS is steadily progressing towards a solid and dependable OS, with features aproaching those of CP/M (a low-end multiuser OS that predates DOS), but with the advantage of greater usability and the ability to run on better hardware than CP/M ever did. Win2003 is now a fully multi-user operating system, unlike 2000 or NT4. I believe that new versions of Exchange actually have a feature that allows you to change the postmaster password - so that companies no longer have to trust their former sysadmins for life (I pray that places I've worked never get hacked - since they can't change that password that will make everyone that has known it a suspect).
As for stability, anyone that has worked with a mixed bunch of servers knows the answer to the MS claims about that. In contrast I've had to reboot a solaris box that is under heavy load twice in the last four months due to hassles with external SCSI hardware - that is the sort of reliability a server OS has to have. Linux can do that too if set up correctly. The uptimes on the 26 linux servers and cluster nodes I look after date back to the last prolonged power failure, or hardware upgrades on some boxes. With windows boxes I've found it better if you reboot them once a week whether they seem to need it or not - solves most of those memory leak problems.
If you consider Canadian SF - The cube didn't have a big budget.
Blakes 7 with good sfx would seem wrong somehow. If were willing to suspend disbeleif enough for FTL travel we may as well suspend disbelief that a red platic esky is a high tech tool kit. The strength of the show was in the characters (well those that were at least 2D), and since it worked well as a radio play recently, lots of expensive visual effects are not required.
I think it stands as one of the few self-consistant SF TV programmes ever made. I also liked the digs at the Trek utopian federation - the trek symbol at 90 degrees and the federation as a police state.
Go back to sleep and let us return to our previously sheduled program - now with 100% more Zen and Orac.
The author of the article has made a common error - confusing strength to weight ratio with strength.
Stength is the amount of force a given cross-section of material can take before it fails. It's given in units of Newtons per millimeter squared (which is a megapascal) or pounds per squre inch. Failure is either when the material is bend permanantly out of shape (yeild strength) or broken (ultimate tensile strength - if it is broken by pulling).
Now a piece of the cheapest steel wire you can can get is going to be stronger than the same cross section of the best bamboo, spider silk or carbon fibre re-inforced plastic you can get. With those light materials you just make it thicker if you want to carry the same load, and you still end up with something ligher than the smaller piece of steel. Those materials have a better strength to weight ratio than steel, but you have thinker peices of material taking the load.
It all comes down to the purpose - you don't want planes or bikes to weigh a lot so you only make the bits that have to be really strong out of steel - say axles and spokes.
That's along the length of the bamboo stems, the direction the fibres run - the bamboo will be strong in that direction under both tension and compression. Under compression it will buckle if you push it hard enough, because it isn't very stiff. The answer is just to use more bamboo.Polymers (eg. cellulose in bamboo) behave differently to metals - their behaviour depends on time as well as load, and something that looks like it has permanantly bent out of shape can sometimes slowly return to its old shape. The change from fairly elastic behaviour (double the force and get double the stretch) to plastic behaviour is hard to spot. Metals behave elasticly, you know Yes, you would just have to design the bike based on the strength of the joint. The same applies with metal bikes, you have to design based on the strength of the weld joint.That means that the bamboo bike would have to be designed a different way to a metal bike. You have to design to use the material, as well as choose the material for the design.Your licence money will not only help push Qt forward, it will also, quite literally, feed the trolls.
I really like X - I prefer to even use twm to windows, but the furthur accross platforms apps stretch the better. My job would be very difficult without *nix apps ported to windows (eg. putty (ssh) and vi), and the more apps that can run on embedded systems the better. Having a web server on a photocopier is a very useful thing, and as more apps are ported things will only get better.
Two charges would apply as far as I know - fraud and demanding money with menaces.
Use those unused CPU cycles in the search for disappearing African Uranium!
They keep satelites in orbit.
Brat - He stole my thingy, and my other thing!
Adult - OK Timmy, I'll put on the blanket everything he has, and you can point out what he stole.
Brat - Won't! He stole my thingy!
For me, I'll just be happy with a working Xnest or other solution that lets me do overlays of different bit depths. Must read more about X, and look at code.
As an example of where X is useful on a low end PC: On Friday on the local machine I was running concurrently 3 copies of X (one nested) with three different window managers with apps from 8 different machines with three very different operating systems. The machine was responsive enough that I don't notice any lag at all, but I wouldn't want to use a similar machine for 3D games newer than Quake][. On a usual day I would run no more than 2 copies of X, and two window managers, and when a very old, very expensive but very useful scientific application is upgraded I will only need one. This application will not run on the local machine at better than glacial speed, let alone disk size considerations, and having a top end PC for everyone that needs to use that application would cost a lot, but no more than a few percent of the licencing costs for each new copy that would be needed.
My co-workers who use windows also use a version of X to get their work done, only they can't work in both 8 bit and 32 bit without closing applications and going into the windows control panel. If their windows video driver and hardware supported overlays they wouldn't need to do that. The desktops run the range from pentium 166 with 32MB up to dual processor atholons, and all run at a good pace in X, since the real work is done on a server and the local machine just does the display for the serious applications. Local stuff, of course, isn't done at the same speed, and the local stuff decides the OS (eg. MS if you need powerpoint). XFree86 on windows in not yet at a stage where it would be as useful to them as the commercial X servers, but it is rapidly getting there. XFree86 runs on five of the machines they use every day, and happily feeds them the windows they need.
To sum up - X and MS Windows are very different things, about a third of the people I work with use both at once on the same screen.
You can theretically run a lower resolution in X under Xnest (Xnest -ac -depth 8 :1) but need the right hardware and drivers. Has anyone had any luck using Xnest this way? On what sort of hardware did it work?
It's very, very different. I myself prefer to keep the real screen size as a virtual desktop when I zoom in, so perhaps the developers didn't see what you are asking for as a worthwhile feature. An application (eg. any Loki game) can resize the screen in X quite happily, but the user is limited to stopping and restarting the server with different options.