🚀 go-pugleaf

In article <815793240snz@c4rv.demon.co.uk>
           raj@c4rv.demon.co.uk "Rajesh Varia" writes:

>In article <815782400snz@wookie.demon.co.uk>
>           wookie@wookie.demon.co.uk "John Beardmore" writes:
>
>> The rules for VL bus are complex.
>> 
>> The number of VL cards you can have depends on the bus speed, something like
>> 
>> 40 MHz, 1 VL card
>> 33 MHz, 2 VL cards
>> 25 MHz, 3 VL cards.
>> 
>There is no actual limit,

Are you certain of that ?  Have you seen the spec ?


>limit, generally you get three slots, though I have seen
>designs with two.  If you want bus-mastering then you have to stick to
>33 MHz for 2 bus-master slots (and one slave) or 1 bus-master (2 slave)
>over 33 Mhz.
>
>> Disk controler cards should go in the master slot, you only have one of those
>> so only one controler card per machine.
>> 
>As stated above, two slots can be bus-mastering.

That's not what I was told was in the spec. Wouldn't work on my machine
either. Are you suggesting that you ought to be able to get a VL BUS
machine with a VL EIDE and a VL SCSI card to work ?

It certainly wouldn't on my machine !


>> Some manuacturers sy that the cards should be at the CPU end of the bus, some
>> not, but the whole thing sounds pretty fragile !  PCI is a lot better thought
>> out and documented !
>> 
>PCI better ? you obviously have not tried PnP.

You obviously haven't had to fight VL bus as much as I have. PCI just
seems to plug in and go !  The PCI bus preforms better too. In the long
run I guess that's more important than configuration issues.

-- 
John Beardmore