Re: Triton DMA

Kai Henningsen (kaih@khms.westfalen.de)
30 Nov 1997 15:31:00 +0200


torvalds@transmeta.com (Linus Torvalds) wrote on 30.11.97 in <65qdbl$c70$1@palladium.transmeta.com>:

> Well, SCSI may be more versatile, but I have yet to get a single report
> about problems with IDE because of bad termination.

On the other hand, Master/Slave/What-have-you jumpering for IDE seems to
be a lot less understandable than SCSI IDs ... not to mention
incompatibilities.

> The fact is that neither IDE not SCSI is perfect. IDE wins hands down
> in price and ease of use, while SCSI wins on high-end performance and
> flexibility. Which one you consider to be more important depends on the
> use.

In my experience, there is not a lot of difference in price (sometimes,
none), and SCSI wins hands-down in ease of use.

> drives etc for SCSI without having to buy an extra card. Anybody who
> buys a SCSI CD-ROM these days has too much money and too little sense.

Or has been burned by IDE too often.

> IDE has had a _lot_ of development due to mass market issues. Many
> technical people look at technical specifications rather than at market
> issues, and that's not necessarily the best thing to do.

Technical issues, and $$. What else is there?

I usually choose cheap stuff for my home machine. For example, I still use
a 486.

I won't put an IDE peripheral on this machine for my own use, ever. Bad
enough that I have to live with them at work. (And even there, servers
usually are SCSI - the few with IDE are *painfully* slow.)

True, SCSI isn't perfect. But IDE isn't even in the running, as far as I'm
concerned.

MfG Kai