On Sat, Jun 22, 2002 at 11:41:44PM +0100, Tony Firshman wrote: > > On Sat, 22 Jun 2002 at 11:52:28, Richard Zidlicky wrote: > (ref: <[EMAIL PROTECTED]>) > > > > >On Thu, Jun 20, 2002 at 12:03:03AM +0100, Roy Wood wrote: > >> > >> In message <[EMAIL PROTECTED]>, Richard > >> Zidlicky<[EMAIL PROTECTED]> writes > >> <SNIP> > >> >Absolutely not. If you are building hardware you can't simply provide > >> >the user with an official SMSQ version in EPROM and a patch on a floppy > >> >disk and expect him to apply the patch to the EPROM. > >> No need. All versions of SMSQ/E for the Qxx (which is what we are > >> talking about here - possibly the GoldFire later but that will have > >> flash ROM) are LRESPR'able over the source code on the ROM. That is what > >> I do because I have an early version of the ROM. > > > >it will break when harddisks are accessed in LBA instead of CHS mode. > That is interesting. I thought Tony Tebby had always intended that the > SMSQ code could be LRESPRed.
you can allways lrespr it from floppy. Keeping HD full backwards compatible with current versions of SMSQ would be probably a lot more work, as far as I could judge this is one of the rather messy aspects of SMSQ. I think most people would like to convert to LBA because that allows - disks > 8GB (once a few other bugs are ironed out) - safe HD image exchange with QXL/QPC > I suspect the majority of users are now running with LRESPRed SMSQ. > >Besides, what is the point to require the user to go through additional > >hoops like this? The speed argument mentioned later in this discussion > >is tripple nonsense and the authors of it should know better. > You need to explain the 'triple' nonsense. - the code is in fast ROM and can be copied from there to RAM much faster than from a HD to RAM, iirc SMSQ already does this. - lrespr SMSQ involves a repeated HW initialisation, count another few seconds on that - it is much easier to screw up your HD than your ROM Richard
