[lkml]   [1996]   [Mar]   [21]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
Messages in this thread
SubjectRe: CONFIG_PNP: Please change the name
In message <>,hab writes:
> I think I fully understand Plug and Play.

You may, but it is not clear from your reply.

> It is a very desirable
> goal to be able to plug in a peripeheral device into a bus slot
> and have the system recognize what resources are required and insure
> no conflicts exist. True it allows software to be written to
> automate this configuration. It is still my contention that a well
> designed board will allow the manual overide of these features for
> the power user, who has the right reason to make the override.
> I would call it Power user Overide.

And in what respect does the ISI PnP spec not allow this ?
All boards (except for those required for booting such as hard disk
controllers), come up quiescent, and can be configured to any settings
under software control.

Of course if the software is braindead and does not allow you to do this, you
have a problem, but it is not a hardware problem nor a flaw in the PnP spec.

> I've been writing software
> and designing hardware too many years to assume that I can do
> everything after the hardware is powered up.

Again, I don't understand your point. There is nothing in the PnP spec
to prevent you doing exactly that. Admittedly there will be cards that
won't implement this properly, but this is not a problem with PnP per se.

> So as long as other
> design considerations allow it I will always choose the board that
> will allow manual setting of at least intial resource allocations.
> I fully prefer those that allow me to do power on configuration, either
> automatic or semi as long as they also give me manual options as well.

It's useful at the moment for working round broken BIOSes, broken cards
and broken configuration managers, but once people get their act together
I can't see any reason why you should want jumpers on the cards and numerous
reasons why you should not.

> I believe that the better products will have both features
> and and are well worth the additional cost. My time is way too
> valuable to try to save the last penny on hardware outlay. I also
> prefer those products that provide excellent documentation, prefererably
> machine readable. Right now Plug N Play still seems to be a selling
> feature that is substituted for poor documentation in some off brand
> boards.

Foul !
Again, just because some people [ab]use it as an excuse to offer shoddy
documentation or do not correctly implement it does not make PnP a bad

> In my recent upgrades I ran into poor PnP and good PnP
> implementations. The good implementations were not afraid to give me
> manual options as well.

This is not the mark of a good PnP implementation. There is nothing wrong
with the boards, but with the BIOS and or configuration software - if these
worked, you wouldn't need kludgy hardware workarounds.

> Be careful in sellecting Plug N Play products,
> Especially if you want to mix it with non-plug and play boards on
> platforms that use more than one OS.

Absolutely. This is not a problem with the PnP spec but with poor and
thoughtless implmentations.

> My recomendation to a hardware
> designer who is developing a new product is still please provide some
> manual control to give the power user some flexability to over come
> unforseen requirements.
> PCMCIA is one place where the manual overrides is probably
> one place where the manual override in impractical.
> By the way there are users that I fully recomend plug and
> play to. I made the assumption that the Hardware person who asked
> about pitfalls was truely interested in why some of us had found
> problems with PnP. I did not say that I had abandoned PnP, only
> that I was selective in which PnP devices I bought.
> Hubert Bahr

I understand, but your reply certainly came across to me as implicating the
PnP standards rather than poor and inadequate implementations.

The ISA PnP spec is a very clever and well thought-out design and kudos
should go to the people at Intel who were responsible. Now it's time for
motherboard manufacturers and software vendors to do an equally decent
job on the software side.

I have kernel code that I knocked up quickly to find and activate my
Supra express 288i PnP modem and I can configure it however I like.


Tim Wright, Worldwide Technical Services, | Email:
Sequent Computer Systems Inc., 15450, | or
SW Koll Parkway, Beaverton, Oregon 97006 | Phone: +1-503-578-3822
"Nobody ever said I was charming, they said "Rimmer, you're a git!"" RD VI

 \ /
  Last update: 2005-03-22 13:36    [W:0.067 / U:6.912 seconds]
©2003-2018 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site