even after it has been pointed out to them that, to the outside world, they appear to be a mob of violent, hate filled idiots.
When has anybody accused them of violence?
Anyone with a strong enough political agenda could be called hate-filled by the terms of the opposing side, and if prior in real life protests show anything it seems they see no reason for protests to not be a fun endeavour while proving their point. This by itself does not make them idiots.
Of course within any large group especially one such as anon where there is no real common anything you are bound to have some troublemakers. But it is because the group is so loose that the term anonymous really doesn't mean anything.
Even if someone were to not like anything of what these other anonymous people are doing, they could still if they wished call themselves anonymous and not be wrong in the slightest by doing so.
one must question why an upstanding genius would associate with a mob of violent, hate filled idiots.
Perhaps because both there are other ways of perceiving them (you are the first person I have ever heard call them violent) and associating with anonymous in reality means absolutely nothing, it is completely disjointed.
easy fix, use your regular smallish password and a checksum algo of your choice (md5sum, sha1sum, whatever) and then use the hash of your crappy password as your actual password.
they then have to figure out what algo you used, what key length, and what simple password, or just try to brute force the ridiculously long password the hash creates.
And all black people like watermelon... because generalizations works so well don't they.
they have no real beliefs or convictions that might give them the courage to continue fighting after a few of them are taken down
Again with the generalizations, how do you know that _none_ of them have any beliefs or convictions? You would have to know every person who has ever claimed to be one personally to know such a thing, which is near impossible considering the people claiming to be anonymous span multiple countries.
After they realize they are not anonymous once a subpoena is involved, and they can go to jail, the members of anonymous will find better things to do with their time.
Or you know, if they haven't done anything illegal and are only being targeted for association with the group they could you know.. fight the charges and continue to speak of any political issues they wish to now with a wider audience.
You can't make such silly generalizations about a group that is so diverse that nobody can really know what each ones motivation is.
You are seriously going to sit there with a straight face and tell me that some 14 year old in the name of activism wrote a DDoS program to distribute so that he could get thousands of others to DDoS visa.com? Seriously, according to you, that is exactly what happened.
why does it have to be a 14 year old, why not a 20 year old, 30 year old? you don't have to be a teenager to let these things loose for a cause you support.
Hell if you were a 14 year old when 4chan first started getting popular you are well over 20 by now.
Here is something that makes much more sense: a different person knew that he could herd the tards very simply since they were just filled with teen angst anyway, so he/she writes a DDoS program, distributes, and herds these 4chan kids into doing it. That way, no traces back to the original person, and they 100% get away with whatever they want while it is the thousands of skiddies doing it for him.
You make it sound like this person is manipulating children to do his bidding instead of simply being with the cause themselves. If someone who agrees with some aspect of what anon is doing is a thirty-something programmer, is it manipulation if they were to just enable them to do what they were already going to?
OK for starters the bit about "consistent speed no matter how many users there were" is complete garbage, with ANY radio based system data system.
When you have highly directional beams going to the location in a point to point manner, and the same likewise back, you can have as many users as antennas you have to point at things.
So you can forget about watching TV while you're using the internet.
Analog tv?? i.e. the tv that is being disabled soon'ish and is going to become free space for which stuff like this would be very useful.
This will not interfere with their digital tv signals, they are on whole different bands.
Secondly, if you start using the TV spectrum for data in both directions, you start putting a really strong signal OUT your TV antenna, which despite being on a different frequency to the actual TV channels, it is close enough to swamp the (really weak by several orders of magnitude) TV signal on the next band with the (extremely strong in comparison) outgoing signal.
so the 64mhz abc analog channel is going to interfere with the 226mhz digital channel hey? what about the 119-124mhz channel, there is no digital channels till about 170mhz or so, they are not going to interfere with anything.
Motorola knew apple were suing already, how do you think apple counter-sued so quickly, and then amended it with more complaints later in reaction to the suit motorola put forth.
Nokia tried for years to work with apple in regards to patents (the iphone uses many many of nokias patents mainly to do with 3g and other wireless business), apple went fuck you to nokia thus the patent battle.
Apple is the one playing hardball here, they are doing what any profit driven company does, tries to force the competition out, by any means possible.
The tenant should pay that, not the landlord, and the landlord should then get reimbursed for the house.
absentee landlord mess
You know, it's actually illegal for landlords to pry too much into the renters lives. They essentially lease the property rights away so renters can actually have privacy to what they are doing.
Jack enumerates the alsa devices and then supports them all..the asound.conf isn't even touched.
This is all very much a non-issue. Something tells me you've never actually used it.There is really no reason to support pulse.
When I can rip out pulseaudio from my pulse based distro, and replace everything with jack in a matter of minutes, how hard do you think configuration of the jack server is?
It does everything pulse does with lower latency and higher accuracy, is more stable and is the only suitable server for pro audio work.
Pulse tends to work for people out of the box (bugs excluded).. until you try to do something it just can't do, then you are boned and have to switch to jack. It's creation was an effort in futility.
Do you even know what Xlib provides in the way of drawing? 1980's-style graphics primitives, pixel-based, non-anti-aliased polylines, circles and arcs.
Now most people these days use X to draw windows and for pointer/mouse events, these can be accelerated using opengl on the graphics card and efforts are under way to do just that, so it isn't just the glx contexts within windows that is 3d accelerated.
You say in another post that "the fundamentals of drawing haven't changed" - Yes, they have.
Fine, you want to get into the fundamentals of 2d/3d, I'll start with geometry, a 4*4 matrix can represent any set of transformations you wish to a set of points in 3d space, you are saying this has changed?.
To avoid gimbal lock such as you get with matrices and Euler angles (which are useful for the lower storage space required) we use quaternions in such situations to store rotational values. Which handily enough is useful for spherical linear interpolation of the rotation. Do you honestly think any of this has changed or is new?
ok enough of that.. I can continue if you want, but long story short none of any of what we see in 3d land is new, we just have more processing power to do more of the same for less money (with fragment programs etc you can implement your own pipeline for different lighting models etc but even all of this is old hat only now we have the processing power to do it).
3D stuff, which is device-dependent
Bullshit, 3d is not device-dependent, there are such things called software renderers, all the math is known and can be done in software. (as an aside, in modern video cards it is all done in software on the video card even, just it has smp to the wazoo)
or you should be doing device-independent rendering.
ok, by some means I have an opengl context, even if it is software, and I have data that is in some kind of vector form (say.. vector fonts) how is this not device independent? and how is this not doing 3d which you say is device-dependent?
Bitmap fonts are the exception, not the rule.
Where did I ever talk about that?
It's insane to expect people to write separate drawing routines for printing, or generating a PDF or whatever. (And X of course never had any kind of real printing support to begin with)
since when does a display server have to care about printing? printing is postscript sent to the printer driver that does as it wishes to get the printer to print the damn thing.
If you are referring to how quartz' internal imaging model strongly correlates to that of the pdf object model, that still does not make it pdf, conversion is needed even if having a similar object model makes creation of the conversion easier. Good point to note is how very strong the postscript and pdf support is in OSS land, this is because for anything vector we have our own conversion bits.
If you think we need X for drawing, then you simply have no clue.
If you think a system that handles pointer events and drawing windows is not useful, you have no clue. Within those windows you can use glx to draw whatever the hell you want with your 3d card (or software renderer).
Learning programming as a way of achieving and end goal sounds like some bad code about to happen.
Necessity is the mother of invention.
Also, some people take up coding because they like what it enables them to do in their spare time (read: implement their own ideas)
The idea and the marketing are what makes the product successful.
An idea without a practical implementation is meaningless. "I have an idea, lets make a cheap flying car" is useless without the engineering (people who implement) being able to actually make it.
and ten years ago like 60% of the browser market used IE6, your point? things change, they only change if we recognize the flaws with the existing system.
The known protocol is decades old, the people that originally wrote them is retired or dead or too old to work hard on it full time like a 20-30something.
FTP is decades old, and the original spec developers are likely reaching retiring age, same with tcp/ip v4, so should we drop all of that too? this argument is silly.
2)This code is so complex, and very very difficult to maintain.
With a project of this size and with this much use, there is constant adaptation going on to new circumstances (3d graphics cards etc) after the new functionality is stable it all tends to get cleaned up. X of today is not the X of 20 years ago, maintainers aren't masochists they do do cleanups etc.
3)The X system is not designed for current desktops, 3d graphic cards, low latency and personal computers, it is designed for mainframes, corporate, boring stuff, making Linux obsolete compared to MacOSX, iOS, or Windows 7 that redesigned their graphic systems.
Bullshit, the fundamentals of drawing have not changed at all, whether 3d acceleration is used for rendering or not is an implementation detail not a flaw with the protocol itself. If you can explain how the fundamentals of drawing things to a screen are different for professional use as opposed to home use, I'd like to hear it.
Again by your logic everything should be reimplemented every five years just because even if it works perfectly reinventing it would make it 'more awesome' this is idiotic.
And AFAIK Jack uses ALSA as output when outputting to the local audio card?
It does so in the same way Pulseaudio does, hint: all actual audio drivers in the kernel are alsa drivers, pulseaudio outputs to alsa in the same way jack does. Alsa is more an abstracted driver interface than anything thus why there is no mixing in kernel.
Configuration with alsa is generally only a hassle when you are using alsa directly with programs because it has no mixing (and the dmix module is questionable), jack and pulseaudio do the mixing then feed the stream to the card through the alsa driver.
Well bsnes is similar in that regard (brings dual cores down to their knees for a snes emu), even when well coded emulation is always a trade-off between accuracy and speed.
That javascript emu is nowhere near accurate and I'd expect a 2-3x hit minimum to bring it up to par with something like VBA, but as you mention in this instance there are other things to worry about in implementation too.
I guess the point I was trying to make is layers of abstraction are there to help us more efficiently code time wise, but when you get layers upon layers upon layers sometimes the benefits just aren't really there and it can help to take a step back.
I guess the kids I come across these days that are 'EVERYTHING IN JAVASCRIPT OR DIE!!' have just hit a nerve with me to an extent. will not surprise me when these same people start wanting raw i/o access to write drivers in javascript too *sigh*
And they would be reimplementing large portions of X's job by doing so. So instead of a known common protocol that is consistent with a few implementation problems, you have a whole new untested drawing system that is GTK specific too... great.
also, a modern system at 1ghz? come off it, 1ghz systems were around in 2002'ish, by 2005 we had 3.2ghz systems (still with a higher ipc count than the 386 mind you). Then we went to slower clocked systems (even then slowest about 1.5ghz in 2006) that completed even more than the faster cycled old ones then upping the clock speed again.
Clock cycles aren't comparable between arch's, but then again I am also taking the grandparents word that those kinds of numbers were being pulled as I wasn't even getting full frames on a quad core i7 with over 10% usage.
Taking that logic to it's extreme means it is fine with you if a calculator app takes multiple cores maxed. After all it is doing something useful isn't it?
There comes a point where ease/speed of development, readability, maintainability, stability, and security trump raw CPU usage.
To an extent I agree, but a gb emu is a matter of a few hundred kilobytes whereas a full on browser requirement is at least several megs if not in the 20'ish range. So far as readability and maintainability you'd probably have an argument for a managed vm'd language like c#.net or java over c or c++ and you'd sacrifice a little speed there too, but javascript? in a browser? seriously?
Call me crazy, but I don't consider running a scripting language in a virtual machine that was basically hacked together any kind of efficient.
It's like demanding the java vm only worse because you really have no idea how well it will perform even on good hardware all depending on the browser.
Yes there is a tradeoff, there always is, but when you're trading 1000% speed difference for 10% difference in development time, with worse code maintenance, I'm free to not use your comparatively slower code and choose the code that had more development time put into it.
The main point is that even if you run wayland, you will still have to run X on top of it.
Wayland has no drawing api, and it's scope is extremely limited compared to x, x will still be needed on top of it for the forseeable future.
I have no idea why there are all these stories that are implying wayland is more than what it is. It sasy specifically on the website that it is not a replacement for x and will need something like X to draw on it *sigh*
depends on your browser son, try and get full frames with firefox 3.6.12 on a core i7, good luck.
chrome and ff4 is a shite site quicker (full frames), but comparatively 1-3% is still ridiculously high considering you can emulate the thing on a 12mhz 386 at full speed without issue.
Slashdot Mirror
even after it has been pointed out to them that, to the outside world, they appear to be a mob of violent, hate filled idiots.
When has anybody accused them of violence?
Anyone with a strong enough political agenda could be called hate-filled by the terms of the opposing side, and if prior in real life protests show anything it seems they see no reason for protests to not be a fun endeavour while proving their point. This by itself does not make them idiots.
Of course within any large group especially one such as anon where there is no real common anything you are bound to have some troublemakers. But it is because the group is so loose that the term anonymous really doesn't mean anything.
Even if someone were to not like anything of what these other anonymous people are doing, they could still if they wished call themselves anonymous and not be wrong in the slightest by doing so.
one must question why an upstanding genius would associate with a mob of violent, hate filled idiots.
Perhaps because both there are other ways of perceiving them (you are the first person I have ever heard call them violent) and associating with anonymous in reality means absolutely nothing, it is completely disjointed.
easy fix, use your regular smallish password and a checksum algo of your choice (md5sum, sha1sum, whatever) and then use the hash of your crappy password as your actual password.
they then have to figure out what algo you used, what key length, and what simple password, or just try to brute force the ridiculously long password the hash creates.
Anonymous are complacent cowards
And all black people like watermelon... because generalizations works so well don't they.
they have no real beliefs or convictions that might give them the courage to continue fighting after a few of them are taken down
Again with the generalizations, how do you know that _none_ of them have any beliefs or convictions? You would have to know every person who has ever claimed to be one personally to know such a thing, which is near impossible considering the people claiming to be anonymous span multiple countries.
After they realize they are not anonymous once a subpoena is involved, and they can go to jail, the members of anonymous will find better things to do with their time.
Or you know, if they haven't done anything illegal and are only being targeted for association with the group they could you know.. fight the charges and continue to speak of any political issues they wish to now with a wider audience.
You can't make such silly generalizations about a group that is so diverse that nobody can really know what each ones motivation is.
You are seriously going to sit there with a straight face and tell me that some 14 year old in the name of activism wrote a DDoS program to distribute so that he could get thousands of others to DDoS visa.com? Seriously, according to you, that is exactly what happened.
why does it have to be a 14 year old, why not a 20 year old, 30 year old? you don't have to be a teenager to let these things loose for a cause you support.
Hell if you were a 14 year old when 4chan first started getting popular you are well over 20 by now.
Here is something that makes much more sense: a different person knew that he could herd the tards very simply since they were just filled with teen angst anyway, so he/she writes a DDoS program, distributes, and herds these 4chan kids into doing it. That way, no traces back to the original person, and they 100% get away with whatever they want while it is the thousands of skiddies doing it for him.
You make it sound like this person is manipulating children to do his bidding instead of simply being with the cause themselves. If someone who agrees with some aspect of what anon is doing is a thirty-something programmer, is it manipulation if they were to just enable them to do what they were already going to?
America creates the most billionaires in the world, and reaps the rewards
And also has a horrible minimum wage in comparison to cost of living.
All that money comes from somewhere, and most of the time it comes from successfully getting away with treating people like dirt.
you know, you can disable most of that stuff
With my instance of kde all I really use is the task bar for switching between applications and the alt-f2 run command.
Nokia smartphone to nokia smartphone, everything transfers just fine.
OK for starters the bit about "consistent speed no matter how many users there were" is complete garbage, with ANY radio based system data system.
When you have highly directional beams going to the location in a point to point manner, and the same likewise back, you can have as many users as antennas you have to point at things.
So you can forget about watching TV while you're using the internet.
Analog tv?? i.e. the tv that is being disabled soon'ish and is going to become free space for which stuff like this would be very useful.
This will not interfere with their digital tv signals, they are on whole different bands.
Secondly, if you start using the TV spectrum for data in both directions, you start putting a really strong signal OUT your TV antenna, which despite being on a different frequency to the actual TV channels, it is close enough to swamp the (really weak by several orders of magnitude) TV signal on the next band with the (extremely strong in comparison) outgoing signal.
so the 64mhz abc analog channel is going to interfere with the 226mhz digital channel hey? what about the 119-124mhz channel, there is no digital channels till about 170mhz or so, they are not going to interfere with anything.
Motorola knew apple were suing already, how do you think apple counter-sued so quickly, and then amended it with more complaints later in reaction to the suit motorola put forth.
Nokia tried for years to work with apple in regards to patents (the iphone uses many many of nokias patents mainly to do with 3g and other wireless business), apple went fuck you to nokia thus the patent battle.
Apple is the one playing hardball here, they are doing what any profit driven company does, tries to force the competition out, by any means possible.
Just VM windows98 or some such on a linux machine for old games, easy as.
The tenant should pay that, not the landlord, and the landlord should then get reimbursed for the house.
absentee landlord mess
You know, it's actually illegal for landlords to pry too much into the renters lives. They essentially lease the property rights away so renters can actually have privacy to what they are doing.
Jack enumerates the alsa devices and then supports them all..the asound.conf isn't even touched.
This is all very much a non-issue. Something tells me you've never actually used it.There is really no reason to support pulse.
When I can rip out pulseaudio from my pulse based distro, and replace everything with jack in a matter of minutes, how hard do you think configuration of the jack server is?
It does everything pulse does with lower latency and higher accuracy, is more stable and is the only suitable server for pro audio work.
Pulse tends to work for people out of the box (bugs excluded).. until you try to do something it just can't do, then you are boned and have to switch to jack. It's creation was an effort in futility.
X has no drawing API!
You then contradict yourself
Do you even know what Xlib provides in the way of drawing? 1980's-style graphics primitives, pixel-based, non-anti-aliased polylines, circles and arcs.
Now most people these days use X to draw windows and for pointer/mouse events, these can be accelerated using opengl on the graphics card and efforts are under way to do just that, so it isn't just the glx contexts within windows that is 3d accelerated.
You say in another post that "the fundamentals of drawing haven't changed" - Yes, they have.
Fine, you want to get into the fundamentals of 2d/3d, I'll start with geometry, a 4*4 matrix can represent any set of transformations you wish to a set of points in 3d space, you are saying this has changed?.
To avoid gimbal lock such as you get with matrices and Euler angles (which are useful for the lower storage space required) we use quaternions in such situations to store rotational values. Which handily enough is useful for spherical linear interpolation of the rotation. Do you honestly think any of this has changed or is new?
ok enough of that.. I can continue if you want, but long story short none of any of what we see in 3d land is new, we just have more processing power to do more of the same for less money (with fragment programs etc you can implement your own pipeline for different lighting models etc but even all of this is old hat only now we have the processing power to do it).
3D stuff, which is device-dependent
Bullshit, 3d is not device-dependent, there are such things called software renderers, all the math is known and can be done in software. (as an aside, in modern video cards it is all done in software on the video card even, just it has smp to the wazoo)
or you should be doing device-independent rendering.
ok, by some means I have an opengl context, even if it is software, and I have data that is in some kind of vector form (say.. vector fonts) how is this not device independent? and how is this not doing 3d which you say is device-dependent?
Bitmap fonts are the exception, not the rule.
Where did I ever talk about that?
It's insane to expect people to write separate drawing routines for printing, or generating a PDF or whatever. (And X of course never had any kind of real printing support to begin with)
since when does a display server have to care about printing? printing is postscript sent to the printer driver that does as it wishes to get the printer to print the damn thing.
If you are referring to how quartz' internal imaging model strongly correlates to that of the pdf object model, that still does not make it pdf, conversion is needed even if having a similar object model makes creation of the conversion easier. Good point to note is how very strong the postscript and pdf support is in OSS land, this is because for anything vector we have our own conversion bits.
If you think we need X for drawing, then you simply have no clue.
If you think a system that handles pointer events and drawing windows is not useful, you have no clue. Within those windows you can use glx to draw whatever the hell you want with your 3d card (or software renderer).
I want this because im not batman
decreasing the likelihood of females wanting to have sex with me.
Something tells me, both of these problems could be solved by simply becoming batman.
Learning programming as a way of achieving and end goal sounds like some bad code about to happen.
Necessity is the mother of invention.
Also, some people take up coding because they like what it enables them to do in their spare time (read: implement their own ideas)
The idea and the marketing are what makes the product successful.
An idea without a practical implementation is meaningless. "I have an idea, lets make a cheap flying car" is useless without the engineering (people who implement) being able to actually make it.
and ten years ago like 60% of the browser market used IE6, your point? things change, they only change if we recognize the flaws with the existing system.
The known protocol is decades old, the people that originally wrote them is retired or dead or too old to work hard on it full time like a 20-30something.
FTP is decades old, and the original spec developers are likely reaching retiring age, same with tcp/ip v4, so should we drop all of that too? this argument is silly.
2)This code is so complex, and very very difficult to maintain.
With a project of this size and with this much use, there is constant adaptation going on to new circumstances (3d graphics cards etc) after the new functionality is stable it all tends to get cleaned up. X of today is not the X of 20 years ago, maintainers aren't masochists they do do cleanups etc.
3)The X system is not designed for current desktops, 3d graphic cards, low latency and personal computers, it is designed for mainframes, corporate, boring stuff, making Linux obsolete compared to MacOSX, iOS, or Windows 7 that redesigned their graphic systems.
Bullshit, the fundamentals of drawing have not changed at all, whether 3d acceleration is used for rendering or not is an implementation detail not a flaw with the protocol itself. If you can explain how the fundamentals of drawing things to a screen are different for professional use as opposed to home use, I'd like to hear it.
Again by your logic everything should be reimplemented every five years just because even if it works perfectly reinventing it would make it 'more awesome' this is idiotic.
And AFAIK Jack uses ALSA as output when outputting to the local audio card?
It does so in the same way Pulseaudio does, hint: all actual audio drivers in the kernel are alsa drivers, pulseaudio outputs to alsa in the same way jack does. Alsa is more an abstracted driver interface than anything thus why there is no mixing in kernel.
Configuration with alsa is generally only a hassle when you are using alsa directly with programs because it has no mixing (and the dmix module is questionable), jack and pulseaudio do the mixing then feed the stream to the card through the alsa driver.
Well bsnes is similar in that regard (brings dual cores down to their knees for a snes emu), even when well coded emulation is always a trade-off between accuracy and speed.
That javascript emu is nowhere near accurate and I'd expect a 2-3x hit minimum to bring it up to par with something like VBA, but as you mention in this instance there are other things to worry about in implementation too.
I guess the point I was trying to make is layers of abstraction are there to help us more efficiently code time wise, but when you get layers upon layers upon layers sometimes the benefits just aren't really there and it can help to take a step back.
I guess the kids I come across these days that are 'EVERYTHING IN JAVASCRIPT OR DIE!!' have just hit a nerve with me to an extent. will not surprise me when these same people start wanting raw i/o access to write drivers in javascript too *sigh*
And they would be reimplementing large portions of X's job by doing so. So instead of a known common protocol that is consistent with a few implementation problems, you have a whole new untested drawing system that is GTK specific too... great.
also, a modern system at 1ghz? come off it, 1ghz systems were around in 2002'ish, by 2005 we had 3.2ghz systems (still with a higher ipc count than the 386 mind you). Then we went to slower clocked systems (even then slowest about 1.5ghz in 2006) that completed even more than the faster cycled old ones then upping the clock speed again.
Clock cycles aren't comparable between arch's, but then again I am also taking the grandparents word that those kinds of numbers were being pulled as I wasn't even getting full frames on a quad core i7 with over 10% usage.
Taking that logic to it's extreme means it is fine with you if a calculator app takes multiple cores maxed. After all it is doing something useful isn't it?
There comes a point where ease/speed of development, readability, maintainability, stability, and security trump raw CPU usage.
To an extent I agree, but a gb emu is a matter of a few hundred kilobytes whereas a full on browser requirement is at least several megs if not in the 20'ish range. So far as readability and maintainability you'd probably have an argument for a managed vm'd language like c# .net or java over c or c++ and you'd sacrifice a little speed there too, but javascript? in a browser? seriously?
Call me crazy, but I don't consider running a scripting language in a virtual machine that was basically hacked together any kind of efficient.
It's like demanding the java vm only worse because you really have no idea how well it will perform even on good hardware all depending on the browser.
Yes there is a tradeoff, there always is, but when you're trading 1000% speed difference for 10% difference in development time, with worse code maintenance, I'm free to not use your comparatively slower code and choose the code that had more development time put into it.
count cache, memory speed and instructions per cycle. We have ridiculously high IPC compared to what we did in the 386 era.
The main point is that even if you run wayland, you will still have to run X on top of it.
Wayland has no drawing api, and it's scope is extremely limited compared to x, x will still be needed on top of it for the forseeable future.
I have no idea why there are all these stories that are implying wayland is more than what it is. It sasy specifically on the website that it is not a replacement for x and will need something like X to draw on it *sigh*
depends on your browser son, try and get full frames with firefox 3.6.12 on a core i7, good luck.
chrome and ff4 is a shite site quicker (full frames), but comparatively 1-3% is still ridiculously high considering you can emulate the thing on a 12mhz 386 at full speed without issue.