But if you do, you're screwed. I've got at least two peecee cards with soft config and no idea where to find the soft config program even if there were no other obstacles to my using it - and the cards don't Just Work, so either they're busted or the soft config is set to something unusual and I don't know what. Effectively, the hardware is likely bricked for me thanks to soft config. (Hmm, the conspiracy theorist in me suggests this is an Evil Plot to sell more cards....) There _is_ that, it's true. With ISA cards, if you soft-config them to an unusual address, they can be approximately impossible to find if you lose your record of the config. I'd really like it if, in addition to a write-protect jumper, they had a "force default address" jumper to recover from that condition. (The same applies in principle to any bus with a large sparse address space, such as Qbus, but not a bus with a small dense address space, such as PCI. I'm somewhat on the fence as to whether I prefer write-enable jumpers or write-protect jumpers - though if you use something like a switch instead, that's irrelevant.) Now, I don't know enough about your particular case to know whether your config includes enough addressing information to produce this "lost config means can't config back" problem, but, if so, you might consider a second jumper/switch/etc with "force default address" functionality - or perhaps make it a three-way setting covering both (RO, soft addr; RW, soft addr; RW, default addr - I see no particular need for RO default addr). /~\ The ASCII Mouse \ / Ribbon Campaign X Against HTML mouse at rodents-montreal.org / \ Email! 7D C8 61 52 5D E7 2D 39 4E F1 31 3E E8 B3 27 4B

Vincent Slyngstad wrote: I've always preferred dip switches over jumpers. Peace... Sridhar

On Tue, 11 Nov 2008, Sridhar Ayengar wrote: Solder pads don't get knocked off, nor bumped into different positions

From: "Ian King" <IanK at vulcan.com> Sent: Tuesday, November 11, 2008 1:35 PM Hey, me too. Darn thing decided to reformat my system disk because of it, too :-(. Vince

Or you could be like other current software products that say helpful things like, " There has been an error." Or, " Unable to complete this task. Please contact your system administrator." Or, "Error -423586293856238392312867896532983." [In a dialog box with an OK button - what's OK about this error?] You're just trying to be too helpful.... ________________________________________ From: cctalk-bounces at classiccmp.org [cctalk-bounces at classiccmp.org] On Behalf Of Jim Brain [brain at jbrain.com] Sent: Tuesday, November 11, 2008 5:00 PM To: On-Topic and Off-Topic Posts Subject: Re: Plug-n-Play for vintage machines? Or, I could just make each of the cartridges a single task board, and have users either insert and remove them as needed, or make their own cartridge expander with switches to manually enable or disable each cartridge slot. Applications could then display screens stating things like: "I'm sorry, but this application requires 16K of expansion RAM for use. Please turn off the computer, insert a 16K RAM Expander or a 24K expansion cartridge with the upper 8KB turned off, and reload this application" "Could you tell me what UART cartridge you have inserted into the computer? I have no way of obtaining that information from the cartridge itself, so you'll have to enter it in my application." "We can't find the UART cartridge at $de40. Since you've no doubt inserted it, it is possible you've not set the address range to $de40-$de5f. Please turn off the machine, ensure the jumper for $de40 (located inside the cartridge shell) is closed and all other are open, and reload this application" "There's a bug in v1.3 of the expander cart, but I can't tell which version you are using, so please select and I will enable the workaround based on your selection" Jim

Ian King wrote: "Possible parity error; All of your life's work and all backups will be reformatted. Is that OK with you?"

On Nov 15, 2008, at 7:02 PM, Roy J. Tellason wrote: Ugh! I've had that happen a couple of times. Another "fun" one, which was easier to spot, was a jumper that had no little metal clip inside. It felt secure (which was odd since it's usually the little metal clip that gives it some friction on the pins) but upon turning it over, nothing. -Dave -- Dave McGuire Port Charlotte, FL

Wouldn't it be more sensible if the voltage selection circuit was designed so that if a particular jumper was open it always gave a lower output voltage tyhan if that jumper was closed (all other jumpers being set the same). Then an open jumper would not fry anything. But then it's my bitter experience that good design and PC hardware do not go together... This story has also convinced me to carry on checking all voltage outputs with a meter before connecting anything to them. In this case, to check the CPU power voltages _at the pins of the CPU socket_ before plugging said chip in. -tony

Ture... Yes, but SMPSUs do approximate a constant power load, so they draw more current as the input voltage drops, hence the damage on too low an input voltage. I would be suprised if any CPU chip had that property. But yes, too low a voltage could case things like internal contentions if 2 drivers were turned on at once. I don't know if this is likely (I don;t have any modern CPUs...) I also don't know how quickly this would cause damage (would you have time to notice the machine wasn't even attempting the POST and then turn it off before the CPU was damaged)? That goes back, alas, to the very first IBM PC. I've often though the PC schematic should be used as an example of how _not_ to design things... -tony

Here is a technique to consider .. I have some Dallas clock chips (1216E) that are designed to work under a ROM chip. By definition it can't have any side effects on the ROM chip that it is hiding underneath. The Dallas trick is to get the programmer to 'wiggle' the address lines with a known pattern before trying to talk to the clock chip. The sequence is long enough such that it will probably never be encounted as part of normal addressing for code reading the ROM. Once the magic sequence is recognized, the clock chip intercepts reads to the ROM chip, and also interprets further wiggling of the address lines as a write operation. If you use a technique like this you can get by without adding any extra hardware. Mike

On 12 Nov 2008 at 21:54, Tony Duell wrote: One can always wire-wrap jumper headers if one doesn't like the idea of the little jumper blocks falling off. Pretty much as permanent as foil-pad jumpers. Cheers, Chuck