DOS ain't dead

Jack has announced a new DMA driver for DOS.

He also released the sources and removed FreeDOS restrictions.

Download location: http://johnson.tmfc.net/dos/drivers.html

WOW!!!
It is really great! Amazing!

Actually, the new Jack's drivers have been available since as early as 26 May!
If you doubt my words, check the date / time stamp of the file "combootf.ima" at

http://johnson.tmfc.net/dos/file/

Note that its date is 6 August, while the date of "drivers.zip" is 9 August. And 6 August is merely the date of last update! The first public versions of Jack's personal drivers (UDMA, UDVD and XMGR) were so available as early as on 26 May!

If you write the image file you'll notice the old Q-drivers there, with the same date and time stamps and file sizes as the final versions of QHIMEM, QCACHE and QCDROM. They even display the original versions and dates when run. However, the similarities with the old drivers end here! The code (and the rest of the data) of the drivers are those of the new Jack's drivers! Thus QCACHE there is in fact UDMA, QCDROM is in fact UDVD and QHIMEM is in fact XMGR, all "disguised" as the old drivers! If you rename the file "run.me" to "runme.btm" and run it in 4DOS, you will read... the text of the file README.TXT included in "drivers.zip"! All aforementioned files are latest-but-one versions which we'll leave there for at least a couple of days (before replacing the files in the image with their real names and latest versions), so that anyone who doubts my words can verify them! (P.S. The 750 MB UDMA cache was proved working by an user whose PC has 1 GB RAM)

> Jack has announced a new DMA driver for DOS.
> He also released the sources and removed FreeDOS restrictions.
> Download location: http://johnson.tmfc.net/dos/drivers.html

Where does it come from ? Did Johnson have the "deleted" sources ? Or did he merge X-drivers code with "source" code "extracted" by the Russian crackers from Q-drivers ? Or did Jack change the mind ?

---
This is a LOGITECH mouse driver, but some software expect here
the following string:*** This is Copyright 1983 Microsoft ***

> Where does it come from ? Did Johnson have the "deleted" sources ? Or did
> he merge X-drivers code with "source" code "extracted" by the Russian
> crackers from Q-drivers ? Or did Jack change the mind ?

You, former NTOSKRNL and current DOS386, don't even tell us your real name, but want to know everything. Perhaps you are a police commissar?! The full source code of all the 3 drivers is freely available there - what more do you want?!

If you tell us who you really are I may write you the history of Jack's drivers. As the Italians say, "tanto per tanto"

Why you are so interested about their provenance? As you are keen on knowledge take this opportunity in order to learn something new. If this would be above your bend, just use the drivers on your emulated DOS environment and see the results for yourself.

> Where does it come from ? Did Johnson have the "deleted" sources ? Or did
> he merge X-drivers code with "source" code "extracted" by the Russian
> crackers from Q-drivers ? Or did Jack change the mind ?

I have the QHIMEM source available, and a brief look into XMGR.ASM unsurprisingly shows that it is a slightly evolved QHIMEM.

---
MS-DOS forever!

> WOW!!!
> It is really great! Amazing!

Yes, big thanks to Jack Ellis (assuming he truly is okay with using this in FreeDOS).

---
Know your limits.h

UDMA interacts more predictable and stable with BIOS SATA disks support! And XMGR hasn't that BX-reg bug. It is nice that this group of DOS drivers is developing again :) I haven't tested UDVD perfomance, but CD-ROM caching sounds promising.

> Where does it come from ? Did Johnson have the "deleted" sources ? Or did
> he merge X-drivers code with "source" code "extracted" by the Russian
> crackers from Q-drivers ? Or did Jack change the mind ?
About cracking old Q-drivers: I've only removed the decryptor stub and decrypted string pointers. After it IDA can be applied to Q-driver and listing can be compiled with TASM. But the source code without data structures and names, with absolute offsets is not very useful -- it is very hard to seriously modify it.

My experiments with decryptor were never published and it is nice that they are useless and source code is freely available :)

So, I suppose that some experimental versions of Q-drivers were used. Jack Ellis definetly had it -- at least to add protection stubs to them before releasing Q-drivers.

P.S. There is no source code of SHCDX33C on Johnson Lam site. I don't know, what is the reason (probably, he and Jack Ellis really erased it).

> P.S. There is no source code of SHCDX33C on Johnson Lam site. I don't
> know, what is the reason (probably, he and Jack Ellis really erased it).

I'm sure it's somewhere, but just use the official version:

http://www.geocities.com/jadoxa/shsucdx/

EDIT: SHCDX33C is still online (must be blind, I completely overlooked it):

http://johnson.tmfc.net/dos/shsucdx.html

---
Know your limits.h

> you are a police commissar

NO.