🚀 go-pugleaf

Alastair Wethers <al@moondog.demon.co.uk> wrote:

>In article <7PsxwGAnIUi0Ew+p@siii.demon.co.uk>, Simon Barfield

>>        there's a problem with *some* seagate hd's in that they need 5v
>>to run and the tx board only supplies 3.3v (or something) on the ide
>>pinouts..(intel changed it without telling anybody)..

Apparently component vendors were informed, but as a system builder I
heard nothing about it until it hit usenet.  Stealth trial (like
FAT32) is my guess; if it works, it will be lauded as a new PII
feature (440LX etc.) and if it doesn't, another FUD bug for socket 7.

>        The problem is not with the general running of the hd, as any
>make and model will run perfectly ok with the tx chipset.  The problem
>appears when formating the disk.  

Now that I can't see, bearing in mind that:

1)   E/UIDE drives are low-level-formatted at factory, not on mobo

2)   OS-level format just writes data to sectors, like any other write

So I can't see how a problem would be miraculously restricted to
OS-level formatting but be totally safe for any other sustained disk
writes, unless it's that the "Format" call writes bigger blocks
causing slightly larger tendency to voltage-drop from load.

Even if the latter is true, that's too small a "safety margin" for me.

Reports I've heard is that affected drives may not be visible at all,
even at the BIOS auto-detect stage.  

They would be the lucky ones; what scares me is the prospect of a
drive that looks like it works but loses data months later...




>-------------- ----- ---- --- -- - -  -    -
When no-one is innocent, everyone is paranoid