Author: Robert Hyatt
Date: 12:36:50 05/25/99
Go up one level in this thread
On May 25, 1999 at 10:37:03, Robert Pope wrote: >On May 25, 1999 at 09:45:16, Robert Hyatt wrote: > >>On May 25, 1999 at 03:06:34, Dave Gomboc wrote: >> >>>On May 24, 1999 at 22:05:17, Dann Corbit wrote: >>> >>>>On May 24, 1999 at 21:00:56, Roger D Davis wrote: >>>> >>>>>Seems that some people think installation will be a bear and others think it's a >>>>>minor issue. It occurred to me that maybe the DB chip could be put into a box >>>>>about the size of my Iomega Zip drive, or just a little larger, and plugged into >>>>>the USB port. Then there would no hardware installation problem, and the thing >>>>>would be portable, too. So carry it back and forth to your local chess club, >>>>>maybe even run it on your laptop on an airplane. Take it wherever. Seems to me >>>>>that would solve a lot of problems, if it could be done. >>>>> >>>>>And that gives it another advantage over the micros: You can't often legally >>>>>install software on multiple computers for easy portability. With a DB box, all >>>>>you'd need to install would be the basic software needed to interface with the >>>>>box, just like Iomega the Iomega Tools disk, for example. >>>>> >>>>>Possible? >>>>I think the only problem with this scheme is that you would have to have all the >>>>memory for the chip included in the box as well. Seems that it might be >>>>expensive to do it that way. The parallel port would be *way* too slow to store >>>>the data to and from memory. Basically, you would have to have a PC in the >>>>little box, along with the Hsu chips, I think. That is clearly doable, since >>>>they have PC-on-a-board implementations, but it would be a lot more expensive >>>>than a card with a chip on it that uses the PC resources. >>> >>>It's clear that the parallel port would be too slow, but he suggested using the >>>USB port. I doubt this would be fast enough either, but I am not informed: I >>>don't really know how fast USB is. >>> >>>Dave >> >> >>A couple of points. USB/Parallel are too slow. This is about high-speed >>communication between the processor and the chess processor(s). It was >>originally built around VME bus speeds. It needs the PCI speed at least. >> >>Second, I don't understand this hesitation to install a PC circuit board. >>Comp-USA does a huge business here in Birmingham. My wife knows _nothing_ >>about computer hardware, yet she has installed a fax modem and an ethernet >>board in her office computer with no help. _students_ are putting machines >>together, much less installing a single board. This is simply a non-issue. >> >>You remove the cover (2-5 screws), remove the blocking plate and throw it away >>(1 screw), insert the PCI board and use the blocking plate screw to hold it in >>the slot, replace the cover, and this is done. It does _not_ require a high >>degree of technical skill. If someone can't follow a simple set of pictures >>and instructions to accomplish this, then it is highly unlikely that person >>can operate the computer at all. >> >>It is _trivial_ and _safe_. > >But then you find out the modem doesn't work because it has to be assigned to an >IRQ below 10, which are all filled. So a tech guy tells you to remove COM2 to >make space, but that was the port your mouse used, so now you have no mouse, and >have to get it running... > >Safe? Most likely, if you aren't a static electricity magnet. Trivial? >Usually, but not always. And until it is, there will always be people nervous >about messing something up. I don't worry about IRQs. :) Linux can handle one device per IRQ, or 10 devices per IRQ with absolutely no problems. So we _never_ have to disable something, or move jumpers, or anything else. One day windows will be as good? :) Bob
This page took 0 seconds to execute
Last modified: Thu, 15 Apr 21 08:11:13 -0700
Current Computer Chess Club Forums at Talkchess. This site by Sean Mintz.