That's strange - I bought the kite from a reputable shop, and flew it "out of the bag" - I didn't change the bridle lines, and although they do tend to tangle, I was very careful to untangle them before flight. Even in strong, steady wind, the pokite would weave back and forth violently. I think it was flying in a stalled condition (in the same way that the classic diamond kite does)
I had exactly the same problem - a physics project designed to lift about 200g of sensor to 200m to measure the "Astronomical Seeing" (in ideally low wind!).
Anyway, our first attempt (wuth dummy load!) was with a Parafoil, sparless 'Pokite' of 2m span. Despite the advice of the kite shop, this was a disaster. It's uncontrollable, oscillates wildly from side to side, flies at a very low angle (25degrees?) on the line, so is poor for gaining height, and worst of all, if the wind ever drops, the kite folds up like a plastic bag with a brick in it, never to recover its unstalled condition.
The second attempt was with a "Megadelta" 3m span Delta kite (with spars). This is perfect for the job: it is very very stable, flies in extremely low wind, stay at an angle of about 75 degrees or higher, and will descend very gently if the wind fails. In fact, in a total flat calm, one can gently reel in the kite on the string, keeping enough forward speed that the kite can be gently flown back to my hand! Liftoff is possible by running, even in windless conditions. (The only problem: in strong winds, it can take 2 people to get it down - the string is rated 1 kNewton for a reason!) We had much observing success with the kite, and I'd highly recommend it. The kite concerned is made by "Skybums" - and there is a photo here (not my website): http://www.nitrobug.freeserve.co.uk/del ta.htm
Sorry, perhaps that wasn't clear. What will probably happen is that the project forks into 2. The PD original is quickly GPL'd and then improved. The PD original will bitrot, because all the developers move to the GPL version. But the government is then left carrying the dead, P.D. fork.
As to why the gov. should prevent MS from making proprietary derivatives: it's because MS then would be able to obtain lock-in. Also, government software should remain open - so that any citizen can continue to participate.
But you then get a fork. The P.D. version gets rapidly abandoned, in favour of someone else's GPL fork, which doesn't really suit the govenment. Also MS could still ship GPL software with windows (like a "distribution") if they wanted.
There's also a pragmatic reason to favour GPL.
Many OSS developers prefer GPL to BSD, and therefore you are more likely to get fixes if you release under GPL.
Besdies which, it's a myth that GPL harms "commercial developers" - indeed they cannot take modifications closed, but they can certainly benefit from the use of the product, and they can even sell it.
/usr/local has its uses. It's for packages installed by the user, which have nothing to do with the distro.
Eg, my distribution comes with Mozilla. (/usr/bin/mozilla). This is used by eg Galeon, so I don't want to break or uninstall it. However, I do want the latest version of mozilla. The answer is to put it in/usr/local/bin/mozilla. (This is even more important with libraries).
Incidentally, Mandrake 10 does have/usr/local/bin in $PATH by default.
Erm - the ipod (as supplied by Apple) doesn't have the horsepower to decode mp3s properly either! I got a 2nd gen ipod, and it mangled mp3s of classical music at 256kbps. After a long investigation, including trying a demo model in the shop, and a 2 month wrangle with Apple's ghastly tech support, I returned it for a refund. But I'd buy another one if I can have Linux/Ogg on it.
The problem is that some of the applications which link to XFree's libraries also have their own licenses. There is now a clash between them, which makes it legally impossible to distribute the applications with the new XFree license. Hence, even if the intentions of both sides are good, it causes a legal mess.
(Incidentally, XFree isn't GPL, but the GPL does do a great job in protecting developers - I know firsthand.)
Re:ok time to start out with first post trolling
on
Is Windows Worth $45?
·
· Score: 1
Mandrake 10.0 is now out - join the Mandrake club and download it, because that way, Mdk get all the money which they would otherwise have to spend on postage,packaging and manuals.
Not quite non-pc, but anyway, I've embedded some of the guts of a 9 button USB mouse (5 button + 2 scrollwheels) into the IBM ultranav keyboard. Using a CMOS 4066 analog switch and a 2 wire touch sensor, I've put a scroll button where the touchpad was. Then I set up XF86 to emulate wheel on the 4th button. So I now have a 3 button trackpoint, with a scroll button too. Neat!
In other words, by removing the synaptics touchpad (ugh!) and adding an extra button, I've managed to revert IBM's new daft mouse button layout to something effectively the same as the original trackpoint layout. (The new layout is daft, because the 3 buttons are in a line, rather than one underneath the other 2. This means that you can't emulate the 3rd button by pressing 1+3, leaving the middle button purely for scrolling.)
Why not choose PHP. It's a very nice language with clear syntax, no need for a compiler, can even be run on Windows, is perfect for the web, and also for scripting. It also has enough in common with C/perl to make it easier to learn these languages (whereas say VBscript is completely foreign). It also has an easy learning curve - it's looseley typed, and gives helpful error messages.
I can recommend the book: SAMS "Teach yourself PHP in 24 hours"
I have one of the UltraNav keyboards - and it is really very good. It has a trackpoint in it with 3 buttons, and is USB - it works fine with Linux. They key feedback is excellent, and the keys are in the right place. (Even better, there's no Windows key, and it has avoided the "function lock" key that MS seems to have imposed on most new keyboards.) It also has page-left and page-right buttons next to the Arrow Keys. Not sure what I make of these yet. They do the same as Alt_L+LeftArrow, Alt_L+RightArrow.
Disadvantages:
1)It has one of these ghastly touchpads in it that I am always pressing by accident. Well, it did:-) I opened up the keyboard, and moved the sensor to somewhere inside the keyboard, and put a blanking plate where it was.
2)The trackpoint isn't quite as sensitive as the one on thinkpads, and the tpctl utiltity (to increase sensitivity) doesn't work via USB.
3)IBM have put their 3 buttons in a line (the old design with a wide "middle button" below the adjacent left and right buttons was far superior). This means that the middle button must handle both scroll emulation AND pasting, which it does via tp-scroll. (I can't click the left+right buttons simultaneously with one thumb).
I shall be altering the circuitry soon to add a real scroll button where the touchpad was...
You can see it here: http://www5.pc.ibm.com/uk/products.nsf/$www partnum lookup/_31P8953
Summary: it's still great - because you can keep your hands on the keyboard all the time, even when mousing.
In the particular case (the wavefinder), better programmers than I, who also have access to a windows system, have not succeeded. This is because a driver alone is not sufficient. The program has to do something really complicated, and undocumented. Therefore, it's necessary to make the supplied windows driver work, so that it can be used by the windows program under WINE.
That's strange. I can't really understand why this would be such a worry for the FCC. Surely most programmers would respect the FCC rules. Whereas for those that wanted to do otherwise, and cause chaos, it's far easier and cheaper to buy a relay and an ignition coil and build a cheap spark-gap transmitter - *lots* of spread-spectrum interference.
Actually, network cards, modems and printers aren't really that interesting. There are usually choices (except if it's a laptop), and so you can always just buy something compatible.
The real place this would be useful is in the case of "weird" hardware where it would be wonderful to be able to have WINE work with the driver. For example, the Psion Wavefinder (USB Digital radio receiver) which at the time was $60, and the only way to get digital radio for less than $450. There is often no alternative to "weird" hardware - anothe brand with similar function may not exist.
The simple answer, which may be of most help is this: RHH 7.1 is ancient, and the bug you have is probably fixed by now. Get a modern distro (eg RH 9) and have another go (you can download it for free, or buy the CDs for a few $). Or if you want something just to test, try www.knoppix.net (no install required). 2 year old versions of wine aren't so great either!
I had the same problem of a site written in ASP and MS Access which I inherited. I'm pleased to say that the wonderful asp2php script did most of the hard work of fixing it. We're now happily using Linux,Postgresql and PHP.
See http://asp2php.naken.cc/
Ground-based alternatives to Hubble do exist.
Using optical aperture synthesis, with a baseline of upto 100m, the *resolving power* of the telsecope can be 50 times better than Hubble. But, Hubble is far superior in sensitivity - due to the lack of a glowing, distorting atmosphere.
COAST is the Cambridge Optical Aperture Synthesis Telescope:
http://www.mrao.cam.ac.uk/telescopes/coast/
I had the same problem - my wonderful Thinkpad A22p has a 1600x1200 LCD at 15" (that's 133 dpi) and the default fonts are almost unreadable. This is what's needed - and it will change fonts globally.
1)In XF86Config-4, add the DisplaySize option like this:
Section "Monitor" Identifier "Generic|Generic Laptop Display Panel 1600x1200" VendorName "Generic" ModelName "Unknown" #Sort out tiny fonts - these are width, height in mm DisplaySize 304 228
2)Change the line in/etc/X11/Xresources to
Xft.dpi: 133
where 133 is the value of xdpyinfo | grep resolution.
Then, restart X and the xfont server (xfs), and log back into KDE. The fonts should all look better (and larger). Hope that helps.
I'd like to propose another googlehole: scientific research. Scientific publishing has a pernicious problem:
1)If you want recognition, and dissemination, you must publish in a recognised, well-known journal. 2)If you want to publish in that journal, you may not publish the same paper elsewhere. (Including the web). You have to give them the copyright. 3)If you want access to the journal, you must have a subscription to it. [N.B. This applies to most of the prestigious journals; but not to all.]
The result is that scientific research results are restricted. Amateurs, and less well funded institutions are locked out.
Even if you have access, (I have the fortune to be at Cambridge Univerity - we have all the subscriptions) there's still no decent search mechanism. Google is blinded, and the proprietary "Web of Science" is very slow, and will only give you a journal reference, not an actual link to the paper concerned! You then have to track down the journal, locate the paper yourself, and hope that the access to it is available online! Or use a (physical) library.
This means that searches are very inefficient (only the abstracts are indexed), and that they are 2 orders of magnitudes slower.
Changing the system is very hard, because no one wants to announce their exciting discovery in an unknown (and widely unread) journal!
Heres something that seems so obvious it can't possibly be OK - can it?
What if I tune a DAB (or FM) radio receiver to a normal, commercial station, and then connect it to a pc. Save all files to disk, and at the end of the day, look at the radio playlist and decide what to keep. Result: over a few months, one gets a large library, all of which is free, and presumably legal. Also, especially if its DAB (digital audio broadcast), it's in high quality.
I agree with you about the power mosfets, although not the charge state. The charge state will be nearly the same if you only run for a few seconds; also there's no reason to repeat exactly the same switching pattern every wave: eg (with just 2 batteries) one cycle might be 00,01,11,01,... but a subsequent cycle might be: 00,10,11,10,... thus evening things out.
Also, the load is partly inductive, so helping the low-pass. Furthermore, the load resistance will not be much larger than the battery internal resistance, so adding extra batteries in parallel *will* increase the current that flows.
It will all be terribly non-linear, but a bit of microprocessor control could optimise that quite well.
I probably wasn't quite clear - I was sugesting a much simpler circuit. Something like this:
[ASCII-art -see below]
Where: 1)The relays are of course solid state for fast switching (although 74 Hz might just be possible with mechanical ones)
2)The circuit is extended to have more batteries/relays in parallel
3)The triggering of relays is set to be slightly out of phase, and with slightly different duty cycles. This uses a microcontroller to provide a "staircase wave".
Taking account of the internal resistances and the circuit inductance, it should be possible to get a reasonable aproximate sine wave using a purely digital "amplifier". At least, one would get the right frequency, and get most of the power into the fundamental.
The advantage is that the amplifier is much more efficient, cheap, and doesn't get very hot.
--------
Unfortunately, the broken slashdot lameness filter won't allow my circuit diagram. Here's a description:
A car battery is connected in series with a relay. 8 of these are duplicated in parallel. This is connected across the speaker. The relay coils (or mosfets) are driven by a microcontroller.
Why use Amplifiers?
on
dB Drag Racing
·
· Score: 2, Insightful
Surely, if they're after max volume, then they don't care about distortion. I'd be tempted just to use solid state relays to drive the speakers with a square wave. That saves lots on cost/complexity/power-disspiation, leaving more for the speakers.
(You could also modify the square wave using n batteries in parallel, switched in for slightly different fractions of the waveform - this would get more power into the fundamental frequency.)
Slashdot Mirror
That's strange - I bought the kite from a reputable shop, and flew it "out of the bag" - I didn't change the bridle lines, and although they do tend to tangle, I was very careful to untangle them before flight. Even in strong, steady wind, the pokite would weave back and forth violently. I think it was flying in a stalled condition (in the same way that the classic diamond kite does)
I had exactly the same problem - a physics project designed to lift about 200g of sensor to 200m to measure the "Astronomical Seeing" (in ideally low wind!).
l ta.htm
Anyway, our first attempt (wuth dummy load!) was with a Parafoil, sparless 'Pokite' of 2m span. Despite the advice of the kite shop, this was a disaster. It's uncontrollable, oscillates wildly from side to side, flies at a very low angle (25degrees?) on the line, so is poor for gaining height, and worst of all, if the wind ever drops, the kite folds up like a plastic bag with a brick in it, never to recover its unstalled condition.
The second attempt was with a "Megadelta" 3m span Delta kite (with spars). This is perfect for the job: it is very very stable, flies in extremely low wind, stay at an angle of about 75 degrees or higher, and will descend very gently if the wind fails. In fact, in a total flat calm, one can gently reel in the kite on the string, keeping enough forward speed that the kite can be gently flown back to my hand! Liftoff is possible by running, even in windless conditions. (The only problem: in strong winds, it can take 2 people to get it down - the string is rated 1 kNewton for a reason!) We had much observing success with the kite, and I'd highly recommend it. The kite concerned is made by "Skybums" - and there is a photo here (not my website):
http://www.nitrobug.freeserve.co.uk/de
Sorry, perhaps that wasn't clear. What will probably happen is that the project forks into 2. The PD original is quickly GPL'd and then improved. The PD original will bitrot, because all the developers move to the GPL version. But the government is then left carrying the dead, P.D. fork. As to why the gov. should prevent MS from making proprietary derivatives: it's because MS then would be able to obtain lock-in. Also, government software should remain open - so that any citizen can continue to participate.
But you then get a fork. The P.D. version gets rapidly abandoned, in favour of someone else's GPL fork, which doesn't really suit the govenment. Also MS could still ship GPL software with windows (like a "distribution") if they wanted.
There's also a pragmatic reason to favour GPL. Many OSS developers prefer GPL to BSD, and therefore you are more likely to get fixes if you release under GPL. Besdies which, it's a myth that GPL harms "commercial developers" - indeed they cannot take modifications closed, but they can certainly benefit from the use of the product, and they can even sell it.
/usr/local has its uses. It's for packages installed by the user, which have nothing to do with the distro. Eg, my distribution comes with Mozilla. (/usr/bin/mozilla). This is used by eg Galeon, so I don't want to break or uninstall it. However, I do want the latest version of mozilla. The answer is to put it in /usr/local/bin/mozilla. (This is even more important with libraries).
Incidentally, Mandrake 10 does have /usr/local/bin in $PATH by default.
Erm - the ipod (as supplied by Apple) doesn't have the horsepower to decode mp3s properly either! I got a 2nd gen ipod, and it mangled mp3s of classical music at 256kbps. After a long investigation, including trying a demo model in the shop, and a 2 month wrangle with Apple's ghastly tech support, I returned it for a refund. But I'd buy another one if I can have Linux/Ogg on it.
The problem is that some of the applications which link to XFree's libraries also have their own licenses. There is now a clash between them, which makes it legally impossible to distribute the applications with the new XFree license. Hence, even if the intentions of both sides are good, it causes a legal mess. (Incidentally, XFree isn't GPL, but the GPL does do a great job in protecting developers - I know firsthand.)
Mandrake 10.0 is now out - join the Mandrake club and download it, because that way, Mdk get all the money which they would otherwise have to spend on postage,packaging and manuals.
Not quite non-pc, but anyway, I've embedded some of the guts of a 9 button USB mouse (5 button + 2 scrollwheels) into the IBM ultranav keyboard. Using a CMOS 4066 analog switch and a 2 wire touch sensor, I've put a scroll button where the touchpad was. Then I set up XF86 to emulate wheel on the 4th button. So I now have a 3 button trackpoint, with a scroll button too. Neat! In other words, by removing the synaptics touchpad (ugh!) and adding an extra button, I've managed to revert IBM's new daft mouse button layout to something effectively the same as the original trackpoint layout. (The new layout is daft, because the 3 buttons are in a line, rather than one underneath the other 2. This means that you can't emulate the 3rd button by pressing 1+3, leaving the middle button purely for scrolling.)
Why not choose PHP. It's a very nice language with clear syntax, no need for a compiler, can even be run on Windows, is perfect for the web, and also for scripting. It also has enough in common with C/perl to make it easier to learn these languages (whereas say VBscript is completely foreign). It also has an easy learning curve - it's looseley typed, and gives helpful error messages.
I can recommend the book: SAMS "Teach yourself PHP in 24 hours"
Richard
I have one of the UltraNav keyboards - and it is really very good. It has a trackpoint in it with 3 buttons, and is USB - it works fine with Linux. They key feedback is excellent, and the keys are in the right place. (Even better, there's no Windows key, and it has avoided the "function lock" key that MS seems to have imposed on most new keyboards.) It also has page-left and page-right buttons next to the Arrow Keys. Not sure what I make of these yet. They do the same as Alt_L+LeftArrow, Alt_L+RightArrow.
:-)
w partnum lookup/_31P8953
Disadvantages:
1)It has one of these ghastly touchpads in it that I am always pressing by accident. Well, it did
I opened up the keyboard, and moved the sensor to somewhere inside the keyboard, and put a blanking plate where it was.
2)The trackpoint isn't quite as sensitive as the one on thinkpads, and the tpctl utiltity (to increase sensitivity) doesn't work via USB.
3)IBM have put their 3 buttons in a line (the old design with a wide "middle button" below the adjacent left and right buttons was far superior). This means that the middle button must handle both scroll emulation AND pasting, which it does via tp-scroll. (I can't click the left+right buttons simultaneously with one thumb).
I shall be altering the circuitry soon to add a real scroll button where the touchpad was...
You can see it here:
http://www5.pc.ibm.com/uk/products.nsf/$ww
Summary: it's still great - because you can keep your hands on the keyboard all the time, even when mousing.
In the particular case (the wavefinder), better programmers than I, who also have access to a windows system, have not succeeded. This is because a driver alone is not sufficient. The program has to do something really complicated, and undocumented. Therefore, it's necessary to make the supplied windows driver work, so that it can be used by the windows program under WINE.
That's strange. I can't really understand why this would be such a worry for the FCC. Surely most programmers would respect the FCC rules. Whereas for those that wanted to do otherwise, and cause chaos, it's far easier and cheaper to buy a relay and an ignition coil and build a cheap spark-gap transmitter - *lots* of spread-spectrum interference.
Actually, network cards, modems and printers aren't really that interesting. There are usually choices (except if it's a laptop), and so you can always just buy something compatible. The real place this would be useful is in the case of "weird" hardware where it would be wonderful to be able to have WINE work with the driver. For example, the Psion Wavefinder (USB Digital radio receiver) which at the time was $60, and the only way to get digital radio for less than $450. There is often no alternative to "weird" hardware - anothe brand with similar function may not exist.
The simple answer, which may be of most help is this: RHH 7.1 is ancient, and the bug you have is probably fixed by now. Get a modern distro (eg RH 9) and have another go (you can download it for free, or buy the CDs for a few $). Or if you want something just to test, try www.knoppix.net (no install required). 2 year old versions of wine aren't so great either!
I had the same problem of a site written in ASP and MS Access which I inherited. I'm pleased to say that the wonderful asp2php script did most of the hard work of fixing it. We're now happily using Linux,Postgresql and PHP. See http://asp2php.naken.cc/
One thing I wonder - why is there so much wasted space on the top half? They could make the screen 40% wider (and still have space for the backlight).
Ground-based alternatives to Hubble do exist. Using optical aperture synthesis, with a baseline of upto 100m, the *resolving power* of the telsecope can be 50 times better than Hubble. But, Hubble is far superior in sensitivity - due to the lack of a glowing, distorting atmosphere. COAST is the Cambridge Optical Aperture Synthesis Telescope: http://www.mrao.cam.ac.uk/telescopes/coast/
I had the same problem - my wonderful Thinkpad A22p has a 1600x1200 LCD at 15" (that's 133 dpi) and the default fonts are almost unreadable. This is what's needed - and it will change fonts globally.
/etc/X11/Xresources to
1)In XF86Config-4, add the DisplaySize option like this:
Section "Monitor"
Identifier "Generic|Generic Laptop Display Panel 1600x1200"
VendorName "Generic"
ModelName "Unknown"
#Sort out tiny fonts - these are width, height in mm
DisplaySize 304 228
2)Change the line in
Xft.dpi: 133
where 133 is the value of xdpyinfo | grep resolution.
Then, restart X and the xfont server (xfs), and log back into KDE. The fonts should all look better (and larger). Hope that helps.
Richard
I'd like to propose another googlehole: scientific research. Scientific publishing has a pernicious problem:
1)If you want recognition, and dissemination, you must publish in a recognised, well-known journal.
2)If you want to publish in that journal, you may not publish the same paper elsewhere. (Including the web). You have to give them the copyright.
3)If you want access to the journal, you must have a subscription to it.
[N.B. This applies to most of the prestigious journals; but not to all.]
The result is that scientific research results are restricted. Amateurs, and less well funded institutions are locked out.
Even if you have access, (I have the fortune to be at Cambridge Univerity - we have all the subscriptions) there's still no decent search mechanism. Google is blinded, and the proprietary "Web of Science" is very slow, and will only give you a journal reference, not an actual link to the paper concerned! You then have to track down the journal, locate the paper yourself, and hope that the access to it is available online! Or use a (physical) library.
This means that searches are very inefficient (only the abstracts are indexed), and that they are 2 orders of magnitudes slower.
Changing the system is very hard, because no one wants to announce their exciting discovery in an unknown (and widely unread) journal!
Heres something that seems so obvious it can't possibly be OK - can it? What if I tune a DAB (or FM) radio receiver to a normal, commercial station, and then connect it to a pc. Save all files to disk, and at the end of the day, look at the radio playlist and decide what to keep. Result: over a few months, one gets a large library, all of which is free, and presumably legal. Also, especially if its DAB (digital audio broadcast), it's in high quality.
I agree with you about the power mosfets, although not the charge state. The charge state will be nearly the same if you only run for a few seconds; also there's no reason to repeat exactly the same switching pattern every wave: eg (with just 2 batteries) one cycle might be
00,01,11,01,...
but a subsequent cycle might be:
00,10,11,10,...
thus evening things out.
Also, the load is partly inductive, so helping the low-pass. Furthermore, the load resistance will not be much larger than the battery internal resistance, so adding extra batteries in parallel *will* increase the current that flows.
It will all be terribly non-linear, but a bit of microprocessor control could optimise that quite well.
I probably wasn't quite clear - I was sugesting a much simpler circuit. Something like this:
[ASCII-art -see below]
Where:
1)The relays are of course solid state for fast switching (although 74 Hz might just be possible with mechanical ones)
2)The circuit is extended to have more batteries/relays in parallel
3)The triggering of relays is set to be slightly out of phase, and with slightly different duty cycles. This uses a microcontroller to provide a "staircase wave".
Taking account of the internal resistances and the circuit inductance, it should be possible to get a reasonable aproximate sine wave using a purely digital "amplifier". At least, one would get the right frequency, and get most of the power into the fundamental.
The advantage is that the amplifier is much more efficient, cheap, and doesn't get very hot.
--------
Unfortunately, the broken slashdot lameness filter won't allow my circuit diagram. Here's a description:
A car battery is connected in series with a relay.
8 of these are duplicated in parallel.
This is connected across the speaker.
The relay coils (or mosfets) are driven by a microcontroller.
Surely, if they're after max volume, then they don't care about distortion. I'd be tempted just to use solid state relays to drive the speakers with a square wave. That saves lots on cost/complexity/power-disspiation, leaving more for the speakers. (You could also modify the square wave using n batteries in parallel, switched in for slightly different fractions of the waveform - this would get more power into the fundamental frequency.)