Time to dump Parallel, Serial PS/2?

anaqer said:
Guden Oden said:
The smaller the fan the higher RPM it needs to do to be of any use and percieved fan noise is first and foremost determined by the RPM
I've never met any 40mm fan that was anything but almost completely silent. I'm doubtful these fans are going to be particulary noisy either; the era when hairdryer fans was l33t is over for good, manufacturers are much more noise-conscious now.

Also, small fans like these are of the kind that craps out much more often than larger ones
Yeah, sure, but that's unrelated to their noise level. :)

My USB devices:

* HP Deskjet printer
* Sony Minidisc cradle
* Logitech MX900 Bluetooth cradle
* MS Sidewinder Force Feedback joystick
* 2x XBox joypads
* MS Intellimouse 3
* Logitech Elite keyboard
* Eizo monitor hub (4-port)
* Eizo monitor controller
* Logitech PS2 headset (used for voice chatting)
* Sony HT-BE1 speaker system (not really used, connected just for fun)
* Lots-in-1 flashcard reader
* Philips webcam (crappy)

I think that's pretty much it. :p
 
My (quite awful) Microsoft USB mouse stays loyal to its PS/2 adapter.

It is incredibly more responsive.

(well, with the PS/2 refresh rate set to 200Hz compared to 125Hz which is maximum on USB)
 
How could it possibly be more responsive when the mouse only REALLY updates as often as the screen refresh? :devilish:
 
Welll, it does 160 - at 320*200...

At that res with AA activated I've noticed Quake2 in OGL mode looks like its software rendered. :D
 
Guden Oden said:
Welll, it does 160 - at 320*200...

At that res with AA activated I've noticed Quake2 in OGL mode looks like its software rendered. :D

My monitor does 120hz at 1024x768, and 100hz at 1600x1200(and 85hz at 1856x1392), but I'm dumping it for an LCD.
 
Guden Oden said:
How could it possibly be more responsive when the mouse only REALLY updates as often as the screen refresh? :devilish:

To get a decent aproximation of the movement you need twice the sampling rate.

I can look out some papers, but not now.

And the USB port is usually a host based, low priority port e.g. if you have high CPU usage your (my) USB mouse will get 1-2 Hz refresh rate. This while the ISA legacy part gets checked every cycle even if you don't have a mouse (or anything else) on it.

Of course these are my observations on my mouse. It could be a problem of actual implementation (i did say it is awful).

PS I run my monitor at 120 hz. And I can make out 85 Hz from 100/120 Hz

anaqer said:
vb said:
[...] compared to 125Hz which is maximum on USB
Not really, no.

Ok. So how can you set it higher?
 
You can pry my 9-year-old PS/2 Logitech TrackMan marble from my dead, cold hands! :devilish:

In other words: If a board doesn't support a PS/2 mouse port, I ain't gonna buy it.
 
I'd say parallel and PS2 mouse are long gone (despite having a parallel printer and that six of the seven functional PC's in this room are using PS2 monitor switcher boxes :D ).

Serial is very close to dead too, the last exclusive use for it (WinDBG, SoftICE, etc.) has finally been replaced by Firewire.

But I don't have a single USB keyboard.
 
incurable said:
You can pry my 9-year-old PS/2 Logitech TrackMan marble from my dead, cold hands! :devilish:

Same goes for my IBM PS/2 buckled spring keyboard. Underside states it's manufactured june 30th 1988 (in the U.K. no less). And it's still better than anything out there.

Cheers
Gubbi
 
vb said:
And the USB port is usually a host based, low priority port e.g. if you have high CPU usage your (my) USB mouse will get 1-2 Hz refresh rate. This while the ISA legacy part gets checked every cycle even if you don't have a mouse (or anything else) on it.
There's obviously something wrong with your setup (too many USB devices, perhaps, or lagging mouse software), Both USB and PS/2 generate an IRQ when something happens and windows does services them as they come (I believe there's no irq scheduling). The only difference I can tell is that the 2nd PS/2 port has an IRQ for its own, instead of sharing it with the rest of USB devices connected to the same header.
 
vb said:
Guden Oden said:
To get a decent aproximation of the movement you need twice the sampling rate.
The nyquist theorem only applies to analog waveforms. The screen is an array of pixels with sharp and finite boundaries, and besides, mouse movement isn't determined by "sampling" anyway. The protocol doesn't work like that.

And the USB port is usually a host based, low priority port e.g. if you have high CPU usage your (my) USB mouse will get 1-2 Hz refresh rate.
Then there's something seriously wrong with your system because I've NEVER seen that happen on a PC with a USB mouse. On the contrary, USB interface sits on the internal chipset bus in modern PCs and is busmaster DMA driven while the PS/2 interface dates back to the very earliest IBM PCs developed in the late 70s or early 80s; it sits on the LPC interface which is nothing other than an 8 or 16-bit 8MHz ISA bus in disguise. Can you say "slow as fuck"? :D

PS/2 mouse interface has very bad latency (from a computer point of view, not neccessarily on a human scale). It's trashy ancient hardware. Kill it dead, now.
 
Gubbi said:
incurable said:
You can pry my 9-year-old PS/2 Logitech TrackMan marble from my dead, cold hands! :devilish:

Same goes for my IBM PS/2 buckled spring keyboard. Underside states it's manufactured june 30th 1988 (in the U.K. no less). And it's still better than anything out there.

Cheers
Gubbi

I also like my Olivetti KDB2812, which has year 1986 on it's type plate. :) This "Made In West Germany" keyboard is not as heavy as original PS/2 keyboard (I have few of them too.) and has slightly worse touch, but this never beeps. You can sit on it and still it happily keeps responsing. (very nice feature if you still play certain games on keyboard.)
 
Back
Top