In message <874sbnawkk.fsf@redmail.netbsd.org>, Chris G. Demetriou writes:
>seebs@plethora.net (Peter Seebach) writes:
>> So, I'm pretty convinced at this point that the killer is that the
>> motherboards I can come up with (ASUS P2L-97, Tyan Thunder 2, both 440LX)
>> are having serious ISA mapping problems.

>That's probably the case.  you might try other cards, to verify that
>they work.  but i'll note that e.g. on my machines that i mentioned
>earlier, i was completely unable to find more than 32KB in one
>contiguous area that could reliably be used for PCMCIA memory.

Well, for the record, IT WORKS!

Solution:  A "via" chipset board.  I got a "supercom" with both Slot 1 and
Socket 370 adapters.

Weakness:  Can't probe my ultra-cheap Trident AGP card, chokes pre-post.
Advantage:  Works with the wireless without any exceptional effort at all.

>I've only seen mention of a couple of these on the web, and invariably
>they've all been back-accessable, so they've not been interestnig to
>me.  8-)

I don't mind back-accessable in this case.  ;)

>> For that matter, anyone know if USB<->CardBus readers are supported in
>> -current?  :)  I know it sounds insane, but it's cheaper than the
>> PCI<->CardBus adapter I've seen.

>I've never seen a CardBus adapter that sits on USB.  I've seen some
>pcmcia storage / cf / smart media adapters, which might work with
>umass, but the one that i didn't present a reasonable device class
>value.  (it was the siig.  *sigh*)

I've seen one in a product catalog.  I thought it seemed crazy.

>> I don't suppose there's an easy way to debug problems with getting access
>> to 64k of ISA memory on a system that *really* doesn't seem to be able to
>> allocate it?

>This is the joy of ISA.  It'll be good when ISA is dead.

Uh-huh.  But then I'll have to get new wireless, and this stuff is fine for me
for now.

-s