it's a case of "I did something unique, and I want to be rewarded for it".
But the whole point of the patent system is -- well, was originally -- that you don't get rewarded for doing something unique. You get rewarded for sharing the special thing that you've done with the rest of the world. Doing something new and unique and then keeping it to yourself is not worthy of reward; which is why if, instead of patenting something you invented, you keep it a secret, then anyone else who gets the same idea independently can patent it in their name -- and you get what people who don't share deserve.
Yes: Continentals generally use a comma to separate the integer from the fraction (for example, "une pile de 4,5 volts"), or sometimes the abbreviation for a measuring unit (for example "1m72 de hauteur"). A comma is easier to write with a fountain pen (which are still in common use on the Continent, where ball-point pens are seen as vulgar) and harder to miss. Even some British engineers and mathematicians write decimal points that look more like commas, which is not a problem because neither engineers nor mathematicians put commas in thousands.
Modern practice is to separate groups of three digits by (en-)spaces. Dots used to be used on the Continent, but this is seen as rather old-fashioned.
Like, for example, the way that the USA made the French invent a word for "inch" (France has used a mètre divided into 100 centimètres since time immemorial) rather than relabel disk drives and monitors being exported to France?
I remember visiting a large independent electrical appliance warehouse and seeing a brand-new computer, running Windows XP; with a terminal emulator (VT320-alike) being used to access a mainframe. That was the only app it was used for. There were older dumb terminals (VT220 clones) around the building; this PC was evidently being used as a replacement for one such.
Companies are paying a fortune in licences where there are Free alternatives available.
I think Ballmer is probably going to start killing stray dogs and cats around Redmond with Open Source Software to prove just how dangerous it is; and in the process, miss out on the patent for an invention which ultimately goes on to symbolise the United States of America around the world.
Heh. I once created a fictitious item to test a stock control system (a Sanderson running PICK) at the last place I worked. The order code I created was CIG-B&H/20 and the description was "Cigarettes, Benson and Hedges, pack of 20". When the system was replaced with a new (but reckoned inferior by everyone who had to use it) Sage Tetra system, we had to get in temps to re-enter everything from the old system into the new one (open source would have made it too easy to create our own import filter). As far as I know, CIG-B&H/20 is still on the stock control system there although there were never any kept in stores.
It depends. You can certainly resolve some sort of signal out of the interference generated by a CRT monitor. Whether or not it bears any resemblance to anything showing on the screen is another matter entirely. This works because of the hefty pulses of current (at audible frequencies) in the horizontal and vertical scan coils, and the fact that electrons are slamming into the screen at high speed. You won't get anything out of an LCD monitor. This uses lower voltage drive signals; and the backlight is usually a cold-cathode tube. This is driven by a high-voltage power supply, which uses a high-frequency oscillator to drive its step-up transformer (the steel core can be much lighter at high frequencies, since the current is flowing in each direction for less time so is less likely to saturate the core). With no requirement for frequency stability or low noise (we only care about getting a high enough voltage to set up an electrical discharge) these power supplies are usually very noisy and tend to mask any other high-frequency, low-strength noise that may be emanating from the monitor.
This works because with UK payphones, all the credit-counting apparatus is in the phone itself, not the exchange. (There is no UK "red box" equivalent -- you've more chance of getting free calls by shining lights down the coin slot.)
Old-style payphones (dial): Dial number, wait for answer, The Pips sound (generated by the phone, triggered by line polarity reversal on connect), insert coin. At intervals the exchange sends a negative voltage pulse down the line, setting off The Pips again. Speaker works during Pips (so you can hear called party announcing their number / name). The short-circuit causes the phone not to see the reversal and believe it is still waiting to be connected.
Newer-style payphones (pushbutton): Insert coin(s), dial number, wait for answer. The Pips sound only after last coin has been used (and modern phones have display screen to show remaining credit). Used to depend on voltage pulses on line; no longer do, as phone knows its own and all local STD codes and keeps track of how much call is costing.
For awhile it was possible to play DTMF tones into a card payphone's mic, then dial 9999 on the keypad (the emergency number is 999 but you need an even number of pulses) and have the phone think it was making a free, emergency call. This was nipped in the bud by filtering out DTMF on all (coin and card) payphone lines, making them respond to LD only; later payphones are DTMF, but with mic muted before connection is established. There was also a payphone with a built-in firmware "cheat" (test mode?) which allowed unlimited free calls to be made if the correct sequence of coins were inserted and the correct code was dialled. These were soon upgraded.....
Any hi-fi audio player has about 20kHz. of bandwidth. CD is restricted by design to 22.05kHz. (to represent a cycle of a waveform, you need to sample at least one crest and one trough). The recording function of portable media players is generally restricted to 8k 8bit samples a second in mono, for 4kHz. of bandwidth. Even medium/long wave radio has potentially 4.5kHz. of bandwidth (station frequencies are always 9kHz. apart, hence every frequency on the MW and LW bands is a multiple of 9kHz).
A video signal contains 25 frames per second, each containing 625 lines. That is 15625 lines a second. Meaning you would have just about enough bandwidth, if you could persuade it to record in hi-fi mono, to record one shade of grey per line; maybe two if you can use both stereo channels. No colour, since the carrier frequency is 4.43MHz. Oh, and you won't have enough bandwidth to fit in any timing information -- the line sync pulses are too short. You might get a frame sync, and your monitor might be able to live with this.
However, lossy audio compression distorts a waveform almost beyond recognition. Your ear might not hear the difference {if the codec was well-chosen} but an oscilloscope trace shows it up clearly. And a TV picture is just an oscilloscope trace, but using brightness to represent the signal amplitude instead of height and spread out over 625 lines instead of all being on one line.
Linux will NEVER, EVER offer a stable binary driver API. The second reason is straightforward ideological purity: one should not have to use any closed-source code in order to make one's computer work. The first reason is that a stable binary driver API is no more or less than a security hole waiting to be reamed with any black hat's cock. Linux aims for compatibility at the source code level; on the basis that having to recompile things just to make them work is better than leaving unpatched security holes just so things can still be used without recompiling.
Of course, the necessity for the user to compile things makes it necessary to release the source code to users. Hardware manufacturers need to get over this. It might take a law to help them.
It's not so much "you're going to Hell when you die", it's more "You won't be able to make any use of any of those files you have saved -- all your letters, all your digital photos, all the music and film clips you've downloaded -- either unless you pay some serious money, or maybe even not at all. Maybe not today, maybe not tomorrow; but soon."
That's so preposterous that most rational people don't want to believe it -- except the ones who have already discovered the hard way that every fucking word of it is true.
"If the truth can be told so as to be understood, it will be believed" -- Terence McKenna.
I think it's very important that "ordinary" computer users are told about the enormous scam that is closed-source software. In no other field of endeavour would the standard practices of the closed-source software industry be tolerated. Whoever heard of a restaurant putting drugs into the food so you would return there again and again, or shoes that would not stay on your feet unless you were also wearing a particular manufacturer's outfits? Yet computer users routinely tolerate vendor lock-in and DRM schemes for ignorance of the alternatives.
We -- the smart techies -- need to explain to non-technical people in simple terms exactly why these things are so bad. And then the smart non-techies will turn away from them. There will always be a few who aren't smart enough to do anything except bend over and take it, but you get that everywhere. We need to explain to people who have been let down by closed-source software how so many of the problems they have suffered arise directly from the closedness of the source, and that there is a better way of doing things.
If we who know the truth don't tell it, the liars will win.
There is a marked difference between GPL and proprietary licences. Nothing in the GPL presumes to interfere with your statutory rights. Many proprietary licences contain clauses which appear to diminish a person's rights under the Law of the Land. In most jurisdictions, that sort of thing is not allowed.
What D-Link were doing was not covered by Fair Dealing Exemption (a right given under the Law of the Land) and so would ordinarily have been considered a violation of copyright law. The German courts have merely ruled that the conditions attached to the permissions afforded by the GPL are conscionable.
In the case of something like the old FrontPage Express licence (with its clause forbidding its use to cast Microsoft in a bad light) or the old Borland compiler licences (with clauses forbidding their use in the development of any product which competed directly with a Borland product) it might be different. Likewise, the restriction on reverse-engineering which appears in many EULAs despite the existence of a statutory right to perform such work in limited circumstances. Such conditions may be ruled unconscionable, if tested in court. So far, they have not been. That in itself is telling.
There is no deterrent effect in destroying your own property to prevent theft. All that happens when people twig onto you doing that, is that your property ends up getting stolen by people whose primary motivation isn't for them to have it for themselves, but for you not to have it. In effect, you're doing the mindless vandals' hard work for them.
Beside which, you may think you have worked hard to earn the money to buy your possessions; but given the way the economy works today, it's almost certain that whoever made them in the first place, worked harder to make them than you did to earn the money to buy them. So someone who steals your hard-earned stuff is just keeping the chain going, really.
What you are basically saying with that attitude is that you don't value the device in its own right so much as you value the fact that you have one and other people don't. I hope you like your own company, because you are unlikely to win many friends thinking like that.
Well, most modern TV sets do have an RGB+SYNC input! Portables may be composite-only; but usually, the AV1 connector of any big-ish set (> 50cm. screen) is wired for RGB and composite. AV2 is usually composite-only or SVHS and composite, and sometimes is shared with the camcorder jacks on the front.
No, HTML tags in capitals make perfect sense when you're making changes to pages on a web server in co-lo using vi (the old-skool BSD one as used in Debian, not vim as in every other distro except the busybox ones) over ssh from an xterm. Think of it as a sort of user-defined, cross-platform syntax highlighting.
Of course, why you'd do it that way, and not use Kate with the fish:// KIOslave (or even run a GUI editor on the server but have it talking to your own desktop's display server), is another matter entirely.
The ARM processor reads (and implicitly decodes) an instruction on the tick; and if the instruction is to be executed (all ARM instructions are conditional), and it doesn't need for an operand to be read from memory, and it involves just one pass through the logic matrix, the result can be ready to write to memory or a register by the tock. If the instruction is not to be executed, but another instruction which is to be executed is right behind it in the pipeline, then the non-executed instruction can be totally ignored -- not even decoded. The pipeline can be long enough to hold both branches of an if... then... else... structure, or even a complete while... { if... then... else... } loop. If all instructions are register-based, the pipeline contents read like a macro..... a user-defined version of the old "CISC instructions executed by an underlying RISC core" thing.
I probably wouldn't have noticed, though, to be honest..... my brain seems to work in floating-point mode! As long as the mantissa is reasonable (it's a small computer, so logically it's bound to measure less than 0.2m. in any direction; but it has room for a CF/MD slot, so it must be at least 0.05m. in one direction) the exponent can be guessed. Ten times too small or too big is an obvious error!
(And to counteract the old chestnut that's bound to crop up sooner or later in any discussion of sensible vs. silly measuring units, it's much easier to divide a metre by six than it is to divide four foot seven by five. A sixth of a metre is 0.167m. to the precision with which anyone can actually read a ruler. But 0.8ft 1.4in. doesn't make sense!)
My assumption is that g is approximately equal to pi * pi. Everybody knows T = 2 * pi * sqrt (l / g). You can cancel the pi with the sqrt(g) to get T ~= 2 * sqrt (l).
If the amp blew up due to a bit of positive feedback, it wasn't very well designed in the first place.
All you are doing when you create a howlround is using an amplifier as an oscillator. At some frequency there will exist a phase shift of an even (odd, if the amp is inverting eor the sound waves are reflected off a hard surface) multiple of pi; and if the loss around the loop is smaller than the gain of the amplifier, then it will oscillate. The amp doesn't know it's in howlround; it's just amplifying some frequency or other. Of course, what with the system being resonant, the signal will be pretty hefty; but an amplifier should be able to handle any signal you throw at it, and throw it out with exactly the same shape, only higher peaks and lower troughs. You can tell a lot from how an amplified square wave looks: if you get overshooting and ringing then there is too much HF response {and a possibility of oscillation at an inaudibly high frequency, perhaps enough to interfere with LW radio}; if you get undershooting and rounding off then there is not enough HF response.
MOSFETs usually have a positive temperature coefficient of resistance: the hotter they get, the less current they admit, and the self-heating effect is reduced. This makes them more proof against catastrophic failure by overheating. Bipolar transistors usually have a negative temperature coefficient. The hotter they get, the more current they admit and the more current they admit, the hotter they get. That's more positive feedback.....
What probably happened was that your amp was a badly-designed bipolar one, and you cooked the output trannies. I'm surprised the fuse on the transformer survived it; most amplifiers' power transformers are woefully underrated.
Ah. What you have to do, you see, is take your piece of string, tie something heavy to the end of it, and set it swinging. Time in seconds the time it takes to swing from one extremity to the other. Multiply that figure by itself. The answer is roughly equal to the length in metres of the string.
All this does is create the possibility for a DOS attack where none existed before.
Beside which, if you are prepared to destroy your own property rather than let it fall into the hands of a thief, you don't deserve to have it in the first place.
Slashdot Mirror
Yes: Continentals generally use a comma to separate the integer from the fraction (for example, "une pile de 4,5 volts"), or sometimes the abbreviation for a measuring unit (for example "1m72 de hauteur"). A comma is easier to write with a fountain pen (which are still in common use on the Continent, where ball-point pens are seen as vulgar) and harder to miss. Even some British engineers and mathematicians write decimal points that look more like commas, which is not a problem because neither engineers nor mathematicians put commas in thousands.
Modern practice is to separate groups of three digits by (en-)spaces. Dots used to be used on the Continent, but this is seen as rather old-fashioned.
Like, for example, the way that the USA made the French invent a word for "inch" (France has used a mètre divided into 100 centimètres since time immemorial) rather than relabel disk drives and monitors being exported to France?
I remember visiting a large independent electrical appliance warehouse and seeing a brand-new computer, running Windows XP; with a terminal emulator (VT320-alike) being used to access a mainframe. That was the only app it was used for. There were older dumb terminals (VT220 clones) around the building; this PC was evidently being used as a replacement for one such.
Companies are paying a fortune in licences where there are Free alternatives available.
I think Ballmer is probably going to start killing stray dogs and cats around Redmond with Open Source Software to prove just how dangerous it is; and in the process, miss out on the patent for an invention which ultimately goes on to symbolise the United States of America around the world.
See here if you don't get it.
Heh. I once created a fictitious item to test a stock control system (a Sanderson running PICK) at the last place I worked. The order code I created was CIG-B&H/20 and the description was "Cigarettes, Benson and Hedges, pack of 20". When the system was replaced with a new (but reckoned inferior by everyone who had to use it) Sage Tetra system, we had to get in temps to re-enter everything from the old system into the new one (open source would have made it too easy to create our own import filter). As far as I know, CIG-B&H/20 is still on the stock control system there although there were never any kept in stores.
It depends. You can certainly resolve some sort of signal out of the interference generated by a CRT monitor. Whether or not it bears any resemblance to anything showing on the screen is another matter entirely. This works because of the hefty pulses of current (at audible frequencies) in the horizontal and vertical scan coils, and the fact that electrons are slamming into the screen at high speed. You won't get anything out of an LCD monitor. This uses lower voltage drive signals; and the backlight is usually a cold-cathode tube. This is driven by a high-voltage power supply, which uses a high-frequency oscillator to drive its step-up transformer (the steel core can be much lighter at high frequencies, since the current is flowing in each direction for less time so is less likely to saturate the core). With no requirement for frequency stability or low noise (we only care about getting a high enough voltage to set up an electrical discharge) these power supplies are usually very noisy and tend to mask any other high-frequency, low-strength noise that may be emanating from the monitor.
This works because with UK payphones, all the credit-counting apparatus is in the phone itself, not the exchange. (There is no UK "red box" equivalent -- you've more chance of getting free calls by shining lights down the coin slot.)
.....
Old-style payphones (dial): Dial number, wait for answer, The Pips sound (generated by the phone, triggered by line polarity reversal on connect), insert coin. At intervals the exchange sends a negative voltage pulse down the line, setting off The Pips again. Speaker works during Pips (so you can hear called party announcing their number / name). The short-circuit causes the phone not to see the reversal and believe it is still waiting to be connected.
Newer-style payphones (pushbutton): Insert coin(s), dial number, wait for answer. The Pips sound only after last coin has been used (and modern phones have display screen to show remaining credit). Used to depend on voltage pulses on line; no longer do, as phone knows its own and all local STD codes and keeps track of how much call is costing.
For awhile it was possible to play DTMF tones into a card payphone's mic, then dial 9999 on the keypad (the emergency number is 999 but you need an even number of pulses) and have the phone think it was making a free, emergency call. This was nipped in the bud by filtering out DTMF on all (coin and card) payphone lines, making them respond to LD only; later payphones are DTMF, but with mic muted before connection is established. There was also a payphone with a built-in firmware "cheat" (test mode?) which allowed unlimited free calls to be made if the correct sequence of coins were inserted and the correct code was dialled. These were soon upgraded
Any hi-fi audio player has about 20kHz. of bandwidth. CD is restricted by design to 22.05kHz. (to represent a cycle of a waveform, you need to sample at least one crest and one trough). The recording function of portable media players is generally restricted to 8k 8bit samples a second in mono, for 4kHz. of bandwidth. Even medium/long wave radio has potentially 4.5kHz. of bandwidth (station frequencies are always 9kHz. apart, hence every frequency on the MW and LW bands is a multiple of 9kHz).
A video signal contains 25 frames per second, each containing 625 lines. That is 15625 lines a second. Meaning you would have just about enough bandwidth, if you could persuade it to record in hi-fi mono, to record one shade of grey per line; maybe two if you can use both stereo channels. No colour, since the carrier frequency is 4.43MHz. Oh, and you won't have enough bandwidth to fit in any timing information -- the line sync pulses are too short. You might get a frame sync, and your monitor might be able to live with this.
However, lossy audio compression distorts a waveform almost beyond recognition. Your ear might not hear the difference {if the codec was well-chosen} but an oscilloscope trace shows it up clearly. And a TV picture is just an oscilloscope trace, but using brightness to represent the signal amplitude instead of height and spread out over 625 lines instead of all being on one line.
You don't have to run IRIX on the Power Mac. You just have to run an X server on the Power Mac.
Linux will NEVER, EVER offer a stable binary driver API. The second reason is straightforward ideological purity: one should not have to use any closed-source code in order to make one's computer work. The first reason is that a stable binary driver API is no more or less than a security hole waiting to be reamed with any black hat's cock. Linux aims for compatibility at the source code level; on the basis that having to recompile things just to make them work is better than leaving unpatched security holes just so things can still be used without recompiling.
Of course, the necessity for the user to compile things makes it necessary to release the source code to users. Hardware manufacturers need to get over this. It might take a law to help them.
It's not so much "you're going to Hell when you die", it's more "You won't be able to make any use of any of those files you have saved -- all your letters, all your digital photos, all the music and film clips you've downloaded -- either unless you pay some serious money, or maybe even not at all. Maybe not today, maybe not tomorrow; but soon."
That's so preposterous that most rational people don't want to believe it -- except the ones who have already discovered the hard way that every fucking word of it is true.
"If the truth can be told so as to be understood, it will be believed" -- Terence McKenna.
I think it's very important that "ordinary" computer users are told about the enormous scam that is closed-source software. In no other field of endeavour would the standard practices of the closed-source software industry be tolerated. Whoever heard of a restaurant putting drugs into the food so you would return there again and again, or shoes that would not stay on your feet unless you were also wearing a particular manufacturer's outfits? Yet computer users routinely tolerate vendor lock-in and DRM schemes for ignorance of the alternatives.
We -- the smart techies -- need to explain to non-technical people in simple terms exactly why these things are so bad. And then the smart non-techies will turn away from them. There will always be a few who aren't smart enough to do anything except bend over and take it, but you get that everywhere. We need to explain to people who have been let down by closed-source software how so many of the problems they have suffered arise directly from the closedness of the source, and that there is a better way of doing things.
If we who know the truth don't tell it, the liars will win.
There is a marked difference between GPL and proprietary licences. Nothing in the GPL presumes to interfere with your statutory rights. Many proprietary licences contain clauses which appear to diminish a person's rights under the Law of the Land. In most jurisdictions, that sort of thing is not allowed.
What D-Link were doing was not covered by Fair Dealing Exemption (a right given under the Law of the Land) and so would ordinarily have been considered a violation of copyright law. The German courts have merely ruled that the conditions attached to the permissions afforded by the GPL are conscionable.
In the case of something like the old FrontPage Express licence (with its clause forbidding its use to cast Microsoft in a bad light) or the old Borland compiler licences (with clauses forbidding their use in the development of any product which competed directly with a Borland product) it might be different. Likewise, the restriction on reverse-engineering which appears in many EULAs despite the existence of a statutory right to perform such work in limited circumstances. Such conditions may be ruled unconscionable, if tested in court. So far, they have not been. That in itself is telling.
There is no deterrent effect in destroying your own property to prevent theft. All that happens when people twig onto you doing that, is that your property ends up getting stolen by people whose primary motivation isn't for them to have it for themselves, but for you not to have it. In effect, you're doing the mindless vandals' hard work for them.
Beside which, you may think you have worked hard to earn the money to buy your possessions; but given the way the economy works today, it's almost certain that whoever made them in the first place, worked harder to make them than you did to earn the money to buy them. So someone who steals your hard-earned stuff is just keeping the chain going, really.
What you are basically saying with that attitude is that you don't value the device in its own right so much as you value the fact that you have one and other people don't. I hope you like your own company, because you are unlikely to win many friends thinking like that.
Well, most modern TV sets do have an RGB+SYNC input! Portables may be composite-only; but usually, the AV1 connector of any big-ish set (> 50cm. screen) is wired for RGB and composite. AV2 is usually composite-only or SVHS and composite, and sometimes is shared with the camcorder jacks on the front.
No, HTML tags in capitals make perfect sense when you're making changes to pages on a web server in co-lo using vi (the old-skool BSD one as used in Debian, not vim as in every other distro except the busybox ones) over ssh from an xterm. Think of it as a sort of user-defined, cross-platform syntax highlighting.
Of course, why you'd do it that way, and not use Kate with the fish:// KIOslave (or even run a GUI editor on the server but have it talking to your own desktop's display server), is another matter entirely.
Gah, what were the odds of me picking some figures at random that made a multiple of five? Oh, right, four to one against. Of course.
The ARM processor reads (and implicitly decodes) an instruction on the tick; and if the instruction is to be executed (all ARM instructions are conditional), and it doesn't need for an operand to be read from memory, and it involves just one pass through the logic matrix, the result can be ready to write to memory or a register by the tock. If the instruction is not to be executed, but another instruction which is to be executed is right behind it in the pipeline, then the non-executed instruction can be totally ignored -- not even decoded. The pipeline can be long enough to hold both branches of an if ... then ... else ... structure, or even a complete while ... { if ... then ... else ... } loop. If all instructions are register-based, the pipeline contents read like a macro ..... a user-defined version of the old "CISC instructions executed by an underlying RISC core" thing.
It's been corrected now -- it says 35 x 105mm.
..... my brain seems to work in floating-point mode! As long as the mantissa is reasonable (it's a small computer, so logically it's bound to measure less than 0.2m. in any direction; but it has room for a CF/MD slot, so it must be at least 0.05m. in one direction) the exponent can be guessed. Ten times too small or too big is an obvious error!
I probably wouldn't have noticed, though, to be honest
(And to counteract the old chestnut that's bound to crop up sooner or later in any discussion of sensible vs. silly measuring units, it's much easier to divide a metre by six than it is to divide four foot seven by five. A sixth of a metre is 0.167m. to the precision with which anyone can actually read a ruler. But 0.8ft 1.4in. doesn't make sense!)
My assumption is that g is approximately equal to pi * pi. Everybody knows T = 2 * pi * sqrt (l / g). You can cancel the pi with the sqrt(g) to get T ~= 2 * sqrt (l).
If the amp blew up due to a bit of positive feedback, it wasn't very well designed in the first place.
.....
All you are doing when you create a howlround is using an amplifier as an oscillator. At some frequency there will exist a phase shift of an even (odd, if the amp is inverting eor the sound waves are reflected off a hard surface) multiple of pi; and if the loss around the loop is smaller than the gain of the amplifier, then it will oscillate. The amp doesn't know it's in howlround; it's just amplifying some frequency or other. Of course, what with the system being resonant, the signal will be pretty hefty; but an amplifier should be able to handle any signal you throw at it, and throw it out with exactly the same shape, only higher peaks and lower troughs. You can tell a lot from how an amplified square wave looks: if you get overshooting and ringing then there is too much HF response {and a possibility of oscillation at an inaudibly high frequency, perhaps enough to interfere with LW radio}; if you get undershooting and rounding off then there is not enough HF response.
MOSFETs usually have a positive temperature coefficient of resistance: the hotter they get, the less current they admit, and the self-heating effect is reduced. This makes them more proof against catastrophic failure by overheating. Bipolar transistors usually have a negative temperature coefficient. The hotter they get, the more current they admit and the more current they admit, the hotter they get. That's more positive feedback
What probably happened was that your amp was a badly-designed bipolar one, and you cooked the output trannies. I'm surprised the fuse on the transformer survived it; most amplifiers' power transformers are woefully underrated.
Ah. What you have to do, you see, is take your piece of string, tie something heavy to the end of it, and set it swinging. Time in seconds the time it takes to swing from one extremity to the other. Multiply that figure by itself. The answer is roughly equal to the length in metres of the string.
All this does is create the possibility for a DOS attack where none existed before.
Beside which, if you are prepared to destroy your own property rather than let it fall into the hands of a thief, you don't deserve to have it in the first place.