Re: SEUL: hardware detection

[William T Wilson <fluffy@dunadan.com>]

On Sun, 18 Jan 1998, Mike Scott wrote:

> thoughts about depending too much on mining info from M$.  It leaves us
> with nothing if the machine doesn't have W95 installed, and M$ might make a

No, it just means that if the machine doesn't have 95 installed, then we
have to either probe for it ourself or the user will have to have
something of a clue.  Since we can count on most of the Clueless(tm)
having Win95 already, I don't think it'll be a problem.  Anybody that
wants to sell a system with SEUL on it will be expected to know what's in
it, and anyone that has managed to dislodge Win95 from their system will
likely know something about it also (except in case of a great data
catastrophe).

> big stink about it (either legally, or simply to publicly sneer "If
> Linux is so much better than Windows, how come they have to steal from

There isn't anything they can do legally.  Microsoft cannot copyright
publicly available information or information which they did not
create (such as monitor lists), nor can they copyright information
generated by their software under direction of the end user (such as
registry information).  It would be comparable to Adobe claiming copyright
over art drawn with Illustrator.

I really hope MS gets to the point where they have to sneer at us.  If a
multibillion dollar company like MS makes such dismal software that a
group of mostly students can for free put together a software package
based entirely on free software that is better than theirs, then who will
be sneering at who?

> Now, mining info like a monitor database is one thing, trying at install
> time to read the config of _this machine_ is another.

Not if we can determine the exact format of the registry... which, I
believe, some of us already have a good handle on.

> That may change, long before SEUL has a chance to become widely

Well, if we can put together a Win32 app that fires up defrag, then runs
fips2 and boots linux, then the presence or absence of FAT32 will not
matter much anyway.

> >with FAT32.  Actually, does anyone know whether DOS-defrag likes to mangle
> >VFAT filenames?
>
> I believe that it does; I wouldn't take the chance.

Well, it's certainly something we can test... of course it's moot if
FreeDOS or whatnot doesn't come with a defrag we can distribute.  We'd
have to either make Win95's do our bidding or write our own, and either
way the long-filename issue goes away.