🚀 go-pugleaf

In article <829351465snz@myth.demon.co.uk>, Mike <zeus@myth.demon.co.uk>
writes

>In article <WZRRxSA31ObxEwne@wookie.demon.co.uk>
>           wookie@wookie.demon.co.uk "John Beardmore" writes:
>
>> You have to look at how cost effective any upgrade is. I've done a whole
>> bunch of CPU upgrades and find that for non speed critical applications
>> like testing new OSs, internet software etc, the upgraded machines,
>> sometimes with very old hardware, can be very cost effective.
>
>You must have different machine requirements to the people I meet.

Probably not. It's just a matter of maximising return on investment...


> The 
>upgrade is usually short lived and can't be moved over, unlike a faster hard 
>disk or video card, memory (depending on existing facilities of course). 

This is not the way it's been here. Give a 386/20 with 16 meg of 30 pin
sims and 700 meg of esdi HD, it seemed to me more sensible to get the
CPU upgraded to a 486/40ish and continue to use the old BIOS, disks and
memory.

I wouldn't want to be buying a fast pentium and putting 1 meg 30 pin
sims on it, or giving it ESDI drives. The best way to preserve the
investment in memory and disk was to give the CPU more grunt.

As things stand, this box with its 130 UKP upgrade CPU makes a fine and
very stable little NT box for comms, turnpike and VA. I have no regrets;
it keeps comms and stuff of the development machines.


>These machines will eventually need a new motherboard anyway.

Not all cases can take a new motherboard; the metal work can preclude
it.

P6 motherboards will probably beed different cases again...


> Considerably 
>cheaper in the medium term.
>
>e.g. Make-It 286 to 486+487        gbp145

Well, 90 anyhow.


>     PCI MB (256k) + Pentium 75MHz gbp170
>
>Such are the bizarre ways of the industry. Horses for courses and all that.

Yup.

-- 
John Beardmore