DOS ain't dead

I'm happy to announce that the new versions of UDMA and UDVD by Jack Ellis now support up to 4 "Legacy" or "Native PCI" IDE controllers and will handle SATA or UltraDMA drives! See http://johnson.tmfc.net/dos/drivers.html

My hex-boot diskette image (http://johnson.tmfc.net/dos/lucho.html) has been updated with them too.

Spread and enjoy!

P.S. Due to real-life issues, I will not be able to follow your comments here...

> I'm happy to announce that the new versions of UDMA and UDVD by Jack Ellis
> now support up to 4 "Legacy" or "Native PCI" IDE controllers and will
> handle SATA or UltraDMA drives! See
> http://johnson.tmfc.net/dos/drivers.html

God bless him!

> My hex-boot diskette image
> (http://johnson.tmfc.net/dos/lucho.html) has been updated
> with them too.

Unsurprisingly.

> Spread and enjoy!
>
> P.S. Due to real-life issues, I will not be able to follow your comments
> here...

Hopefully nothing serious.

---
Know your limits.h

> > Due to real-life issues, I will not be able to follow your comments here...
>
> Hopefully nothing serious.

Thanks, nothing to worry about, but I'm just too busy these days / weeks...

> > I'm happy to announce that the new versions of UDMA and UDVD by Jack
> > Ellis now support up to 4 "Legacy" or "Native PCI" IDE controllers
> > and will handle SATA or UltraDMA drives! See
> > http://johnson.tmfc.net/dos/drivers.html

Meanwhile, newer versions of UDMA and UDVD have been released (dated 23 September 2007), which now run NVidia NForce4 SATA boards properly (i.e. support multiple devices with identical PCI class data).

A huge post-scriptum follows, for which I apologise in advance:

If someone could still post to the Udo's semidead forum (I can't, however I try!) please announce this there too. And of course the FreeDOS mailing lists, although Jack's drivers are not limited to FreeDOS - they support any modern DOS kernel. I am writing from work and unfortunately can't easily fiddle with e-mail here which is required to access their mailing lists. So if you're a subscriber of theirs... Thanks very much for advance. The SATA support is quite a big news for DOS users!

> A huge post-scriptum follows, for which I apologise in advance:

No huger than the (now locked) OpenWatcom thread!

> If someone could still post to the Udo's semidead forum (I can't, however
> I try!) please announce this there too. And of course the FreeDOS mailing
> lists, although Jack's drivers are not limited to FreeDOS - they support
> any modern DOS kernel.

I'm not on any lists, so I just e-mailed Jim, Eric, and Udo directly. That should suffice!

> The SATA support is quite a big news for DOS users!

Any news is big news.

---
Know your limits.h

> I'm not on any lists, so I just e-mailed Jim, Eric, and Udo directly. That
> should suffice!
>
> > The SATA support is quite a big news for DOS users!

Jim Hall has announced it on the main FreeDOS site.

---
Know your limits.h

> > I'm not on any lists, so I just e-mailed Jim, Eric, and Udo directly.
> > That should suffice!

Thank you!

> > > The SATA support is quite a big news for DOS users!
>
> Jim Hall has announced it on the main FreeDOS site.

He announced the latest news that it now supports NVidia's multiple controllers with same PCI class data, and not the previous bigger news that it now supports SATA, but the smart DOS user will notice the magic word "SATA" and understand

Also, the new UDVD completely supersedes the modification of XCDROM by Mark Tsai. But because none of us the 3 beta-testers of Jack possesses a SATA CD-ROM or DVD drive (including the whole CS department of our university!), this is not tested. So, if anyone has this latest scream of the fashion, a test would be appreciated. The code is carefully checked and it should work, but Donald Knuth wrote in 1977: "Beware of bugs in the above code; I have only proved it correct, not tried it."

> But because none of us the 3 beta-testers of Jack possesses a
> SATA CD-ROM or DVD drive (including the whole CS department of our
> university!), this is not tested. So, if anyone has this latest scream of
> the fashion, a test would be appreciated. The code is carefully checked
> and it should work, but Donald Knuth wrote in 1977: "Beware of bugs in the
> above code; I have only proved it correct, not tried it."

Ok, then everyone owning a SATA CD/DVD drive please test the driver and post the result here!

Once there are 2 or 3 positive results I can eventually add the additional init code to the XCDROM JLM driver.

---
MS-DOS forever!

> Ok, then everyone owning a SATA CD/DVD drive please test the driver and
> post the result here!

Why not.

> Once there are 2 or 3 positive results I can eventually add the additional
> init code to the XCDROM JLM driver.

The release of the "free use" XMGR/UDMA/UDVD source code obsoletes XDMA/XCDROM. And in my humble opinion, mixing old and new code is not the best possible idea. Albeit not formally required, a request for permission from Jack for use of his new source code in other programs would greatly be appreciated by him. Ethics!

> Why not.

This wasn't a question.

> The release of the "free use" XMGR/UDMA/UDVD source code obsoletes
> XDMA/XCDROM. And in my humble opinion, mixing old and new code is not the
> best possible idea. Albeit not formally required, a request for permission
> from Jack for use of his new source code in other programs would greatly be
> appreciated by him. Ethics!

In *my* humble opinion it is a bit strange if someone who illegally distributes an MS-DOS "clone" requires ethically thinking from somebody else. As for me, I won't talk to Mr. Ellis. But no need to worry, I won't use his source then ... and this issue has very low priority. However, since the source apparently is NOT free (at least not in the sense I understand freedom), it might be a good idea to clarify the conditions when and how to use it in the readme.txt.

> UDMA/XMGR makes XDMA/XCDROM completely obsolete

No. SATA DVDs are rare, because they offer no benefits compared to PATA - in contrast to SATA HDs. And adding support for SATA is rather simple, since the the ATA/ATAPI port interface is still valid.

---
MS-DOS forever!

> In *my* humble opinion it is a bit strange if someone who illegally
> distributes an MS-DOS "clone" requires ethically thinking from somebody
> else. As for me, I won't talk to Mr. Ellis. But no need to worry, I won't
> use his source then ... and this issue has very low priority. However,
> since the source apparently is NOT free (at least not in the sense I
> understand freedom), it might be a good idea to clarify the conditions
> when and how to use it in the readme.txt.

What "clone" do you mean? LZ-DOS? Have you read that its free to use and convey? As to Jack Ellis, you obviously fear his new source code, which is free, by the way - so if you can't understand it, please find a better excuse not to use it than falsely calling it "non-free"!

> I am not a lawyer, and I have no wish to be one. Anybody is
> "free" to use the source files for the UDMA/UDVD/XMGR drivers
> any way they wish. I hope THAT is clear-enough for you all.

The above was written by him at http://www.mail-archive.com/freedos-user@lists.sourceforge.net/msg06151.html

> > UDMA/XMGR makes XDMA/XCDROM completely obsolete
>
> No. SATA DVDs are rare, because they offer no benefits compared to PATA -
> in contrast to SATA HDs. And adding support for SATA is rather simple,
> since the API is the same.

There are other reasons why XDMA and XCDROM are obsolete, but obviously you don't understand them, or don't care about them, or both, Mr. Andreas Grech.

Ok, another one of those pathetic "debates".

> What "clone" do you mean? LZ-DOS? Have you read that its free to use and
> convey?

LZ-DOS is just a stolen MS-DOS. Period.

> As to Jack Ellis, you obviously fear his new source code,
> which is free, by the way - so if you can't understand it, please
> find a better excuse not to use it than falsely calling it "non-free"!

If I have to get a permission from Mr Ellis to use the source, it cannot be called "free". These "he fears the source code" and "he can't understand it" remarks are from Mr. Ellis, if I understand you correctly. Please excuse me that I'll choose to ignore them then!

> There are other reasons why XDMA and XCDROM are obsolete, but obviously
> you don't understand them, or don't care about them, or both, Mr. Andreas
> Grech.

Dear friend, do you think this type of posts gives you much credit? I have some doubts about that.

---
MS-DOS forever!

> If I have to get permission from Mr Ellis to use the source, it cannot
> be called "free". These "he fears the source code" and "he can't
> understand it" remarks are from Mr. Ellis, if I understand you correctly.
> Please excuse me that I'll choose to ignore them then!

I do not CARE who uses my source files, nor for what purpose. I "Gave
UP" caring, after my files were mercilessly STOLEN by one guy from your
country and two others from the Czech Republic and Russia. If I still
DID care about sources, I would NEVER have released XMGR/UDMA/UDVD with
full source files! But I have better things to do than fight THIEVES!

Lucho's comments to you are NOT AT ALL from me! The idea that you may
"fear" my source code is his idea, NOT mine. I will admit that I told
him and Johnson Lam how UPSET I am, that so many guys in this sad World
can "use" others' software WITHOUT even the common "courtesy" of ASKING
if this is acceptable BEFOREHAND! Just see how QUICKLY my new drivers
"appeared" on the IBiblio server, without ANYBODY at the Circus Maximus
ASKING if it was O.K. with ME! Even AFTER my posting about "lawyers",
I guess they STILL must have believed I might yet say "No". WRONG!!!!

If you also can't understand how obsolete XDMA/XCDROM now are, for want
of many cumulative "bug fixes", caching, SATA support, and possibly for
"automatic" caching in UDVD (it may soon LOSE its /C switch!), then you
need to learn how many BAD BIOS programs still run SATA in DOS systems,
and you need to learn how slow CD/DVD "seeks" are. Caching ALL CD/DVD
directories (not merely the last-used block as in SHCDX33C), eliminates
those "seeks" and gives me 40% faster CD/DVD speed, as it gives others!

If you like using obsolete software, that is your business. But DON'T
"assign" words to ME before you even try to learn the TRUTH! It tends
to make you look like a DAMNED LIAR!!!

> I will admit that I told him and Johnson Lam how UPSET I am, that so
> many guys in this sad World can "use" others' software WITHOUT even
> the common "courtesy" of ASKING if this is acceptable BEFOREHAND!

I don't understand this. Do you refer to various distros publicly redistributing your stuff? (Examples welcome.)

Distribution terms for software vary widely in their so-called requirements. Usually, people just ignore it as long as the software is free and available online. (Not everyone has the time and energy to be 100% diligent as the FSF, for instance.) This is usually not meant to be offensive.

> Just see how QUICKLY my new drivers
> "appeared" on the IBiblio server, without ANYBODY at the Circus Maximus
> ASKING if it was O.K. with ME! Even AFTER my posting about "lawyers",
> I guess they STILL must have believed I might yet say "No". WRONG!!!!

Well, iBiblio is just an official mirror for FreeDOS, so Jim Hall often backs up most everything there (even stuff not officially part of FreeDOS). I mean, it's easy to "lose" software if the official site goes down and nobody mirrored it. That has indeed happened a few times. (Granted, even iBiblio isn't completely up-to-date, but it's much better than nothing, IMO.) He only mirrors what he personally finds useful, though, so obviously he likes and appreciates XMGR/UDVD/UDMA.

What, you don't like backups or having old versions or people possibly getting old versions or what? (Just asking for clarity's sake.)

---
Know your limits.h

> If you like using obsolete software, that is your business. But DON'T
> "assign" words to ME before you even try to learn the TRUTH! It tends
> to make you look like a DAMNED LIAR!!!

There was a remark from Mr. Lucho which gave the impression that he's repeating words coming from you. No big deal IMO, and there's no lie involved. However, better be a bit careful calling other people "liars"! I very well remember "the Q binaries are now orphans without any source files" thingy ( http://johnson.tmfc.net/dos/qcdrom.html ), which made you look ... not "fully honest".

---
MS-DOS forever!

> There was a remark from Mr. Lucho which gave the impression that he's
> repeating words coming from you. No big deal IMO, and there's no lie
> involved.

You assigned my words to Jack just because of your false impressions, thus lying!

> However, better be a bit careful calling other people "liars"! I
> very well remember "the Q binaries are now orphans without any source
> files" thingy ( http://johnson.tmfc.net/dos/qcdrom.html ), which made you
> look ... not "fully honest".

Long before deleting the Q* sources, Jack "forked" a version for his own personal use (which was meant to never leave his computer), in which he removed some code and fixed some bugs. So he was fully honest when he wrote that he had deleted the Q* sources. And his new drivers stem from this "forked" version (or as he called them, "my personal drivers"). Or you deny him the right to "fork" his own code?!

Words like "steal/stolen", "pirate/pirated" and so on are so inappropriate to not only myself but also software in general that I simply refuse to comment on them!

> Long before deleting the Q* sources, Jack "forked" a version for his own
> personal use (which was meant to never leave his computer), in which he
> removed some code and fixed some bugs. So he was fully honest when he
> wrote that he had deleted the Q* sources. And his new drivers stem from
> this "forked" version (or as he called them, "my personal drivers"). Or
> you deny him the right to "fork" his own code?!

Yes, then from a technical point of view it was no lie. So he is safe "de jure", but "de facto" it's not that convincing.

Jack deserves respect because of his hard work for DOS, but you deserve nothing. All you're doing is abuse and destruction.

---
MS-DOS forever!

> > I am not a lawyer, and I have no wish to be one. Anybody is
> > "free" to use the source files for the UDMA/UDVD/XMGR drivers
> > any way they wish. I hope THAT is clear-enough for you all.
>
> The above was written by him at
> http://www.mail-archive.com/freedos-user@lists.sourceforge.net/msg06151.html

You have defeated your own case. If it's free, permission to use it is not required. If Jack Ellis wishes to control use of his code, he should say clearly that it is "not free", and may be used only with his consent.

> There are other reasons why XDMA and XCDROM are obsolete, but obviously
> you don't understand them, or don't care about them, or both, Mr. Andreas
> Grech.

If Japheth or anyone else doesn't appreciate the miracle of the newer drivers, then enlighten them. Insulting people for nothing doesn't open their minds.

> And in my humble opinion, mixing old and new code is not the
> best possible idea.

A strange thing for you to write. Revived 4DOS does not mix code?

The fact is, code gets mixed a lot - what else is any upgrade or revision or library use?

> You assigned my words to Jack just because of your false impressions, thus
> lying!

At worst, it is a mistake, not lying. There is a very large difference. As for impressions, you do not encourage generosity.

> Words like "steal/stolen", "pirate/pirated" and so on are so inappropriate
> to not only myself but also software in general that I simply refuse to
> comment on them!

To say that they are inappropriate as subjects of discussion, usually has one purpose: To give oneself a license to steal.

> If Japheth or anyone else doesn't appreciate the miracle of the newer
> drivers, then enlighten them. Insulting people for nothing does not
> open their minds.

Very well!!

Try a simple test. Run whatever DOS variant you like with NO EMM386,
JEMM386, or anything equivalent, and with NONE of my UltraDMA drivers.
Run whatever set of disk copies/compares you like, and note the speed.

Then, reload your DOS system using exactly the same configuration, and
do exactly the same test again. But THIS time, use "Debug" and patch
location 0:47Bh from 014h to 034h, BEFORE you run your test. I would
bet you "Dollars to Navy-Beans" that by adding the 020h bit at address
0:47Bh, your second test runs 5 or 6 times SLOWER!

What does the 020h bit at 0:47Bh denote? That "Virtual DMA Services"
are active, which many usual "El Cheapo" BIOS programs DO NOT support!
Without VDS, i.e. with no ability to find the "virtual 32-bit address"
of a user I-O buffer, a "cheap" BIOS must use only "PIO mode" for data
transfers! I hope I do not have to tell you how SLOW "PIO mode" is!!

Who must set the 020h bit at address 0:47Bh?? ANY driver that sets a
system in "protected mode" and so demands "virtual 32-bit addressing",
i.e. EMM386/QEMM/386MAX ... and JEMM386!! This is why my README file
for XMGR/UDMA/UDVD no-longer recommends running ANY hard disk in a DOS
system using only BIOS I-O. Too many people DO load "protected mode"
memory managers that require "VDS services", and too many "cheap" BIOS
programs will then RUIN your hard-disk speed!!

About caching, and despite comments by others that SATA disks are fast
enough that they do not NEED caching, I shall only say: Try your OWN
tests of whatever DOS you like, with and without caching. My system
performs TWICE as fast writing files and slightly faster reading them,
when I do load UDMA, even using its 5-Megabyte "minimum" cache. But,
don't just "take my word" -- TRY IT YOURSELVES, before you believe any
BULL**** that SATA disks now do not require caching in DOS!!

As for caching CD/DVD drives, in the face of their miserably SLOW seek
times, do try copying "loaded" 635-MB CD platters (I use an old NVidia
Driver CD) from CD to disk. Just use "XCOPY F:\*.* C:\TEST /S" or an
equal /S command, which copies EVERYTHING from CD to a disk directory.
Note how fast it runs using UDMA/UDVD and caching (usually 125 seconds
for me). Then, try it WITHOUT caching, i.e., use UDVD but do not use
UDMA. 195 seconds for me, and I bet "Dollars to Navy Beans" that you
get a similar LOW speed with your systems. With no caching of CD/DVD
directories, CD/DVD seeks will MURDER you!!

But again, don't just "take my word"!! TRY all this, for YOURSELVES,
especially setting the 020h bit at 0:47Bh!! THAT should prove to you
the kind of total CRAP that most mainboard vendors put in their BIOS!!
After that, if you still wish to believe "False Prophets" re: the need
for UltraDMA drivers and for caching, I have a Bridge in Brooklyn that
might be of interest to you!!

> > And in my humble opinion, mixing old and new code is not the
> > best possible idea.

I mean mixing code from the incompatible UDVD (new) and XCDROM (old).

> A strange thing for you to write. Revived 4DOS does not mix code?

No. The above incompatible mix of existing old and new code doesn't occur there.

> The fact is, code gets mixed a lot - what else is any upgrade or revision
> or library use?

You probably misunderstood me.

> > > I am not a lawyer, and I have no wish to be one. Anybody is
> > > "free" to use the source files for the UDMA/UDVD/XMGR drivers
> > > any way they wish. I hope THAT is clear-enough for you all.
>
> You have defeated your own case. If it's free, permission to use it is not
> required. If Jack Ellis wishes to control use of his code, he should say
> clearly that it is "not free", and may be used only with his consent.

Where did I write that it's required?! It's not! I just wrote that asking for permission would be appreciated! Only the laws of ethics require it, which fortunately have nothing to do with the lawyerish software licences, so beloved by some here. So please read again Jack's words quoted above and don't misinterpret mine, like Japheth does!

> If Japheth or anyone else doesn't appreciate the miracle of the newer
> drivers, then enlighten them. Insulting people for nothing doesn't open
> their minds.

Don't be naive, he appreciates their miracle very much, but his pride will never let him admit that! Thus his reluctance to ask for permission, under the pretext that if it's required to use Jack's code, it becomes non-free. But it's not so!

> At worst, it is a mistake, not lying. There is a very large difference. As
> for impressions, you do not encourage generosity.

But Japheth does this intentionlally, so it's not just a mistake of his.

> To say that they are inappropriate as subjects of discussion, usually has
> one purpose: To give oneself a license to steal.

Software cannot be stolen. It can just be copied, unlike physical objects.

As to my "illegal" activity, only a court can say what's legal and what's not.

> Jack deserves respect because of his hard work for DOS

I'm glad that you admit that!

> but you deserve nothing. All you're doing is abuse and destruction.

Thank you!!! I will remember these words!!! And what have I destroyed, please???

You're a ****. This supersedes by far the fact that you're a good programmer

How come all this needs to lead to gripe each time ?

I believe we need an END to all this, so let us proceed as follows:

Those of you who like XMGR/UDMA/UDVD, use them.

Those who don't like them, me, or both, use whatever else you wish.

Those who are in doubt, do your OWN tests, then decide for yourself
what DOS drivers you wish, without advice from me nor anybody else.

Those with more comments re: XMGR/UDMA/UDVD, please address them to
<http://johnson.tmfc.net/> as before.

Please do not send comments about XMGR/UDMA/UDVD to this nor to any
other boards.

I am damned-well TIRED of all this, yet again!! I am totally SICK
of doing the best job my 42 years of assembly-language allows me to
do in DOS disk/CD/DVD drivers. And my ONLY result is AGAIN having
to deal-with thoughtless, undisciplined, conceited, SPOILED BRATS!!

So, let us save "friend" Riebisch some file space on his server and
END all this!! Use my drivers if you like them. If you don't, I
now have but 6 words for you all: SHUT THE FUCK UP, ABOUT THEM!!!!

> > > And in my humble opinion, mixing old and new code is not the
> > > best possible idea.
>
> I mean mixing code from the incompatible UDVD (new) and XCDROM
> (old).
>
> > A strange thing for you to write. Revived 4DOS does not mix code?
>
> No. The above incompatible mix of existing old and new code doesn't occur
> there.
>
> > The fact is, code gets mixed a lot - what else is any upgrade or
> revision or library use?
>
> You probably misunderstood me.

I didn't misunderstand. You failed to specify "incompatible". But since you have now, I thank you for the lesson in programming. I even plan to extend it to hardware - I'm going to stop using telephone wire in power circuits.

> Thank you!!! I will remember these words!!! And what have I destroyed,
> please???

Well, one example are your little attempts to ruin FreeDOS's reputation, my big hero Lucho (http://www.drdosprojects.de/forum/drp_forum/posts/4363.html).

A small hint, my friend: unlike the "Evil DoctoR" board you won't be able to hide behind the back of the admin here.

---
MS-DOS forever!

> Use my drivers if you like them. If you don't, I
> now have but 6 words for you all: SHUT THE FUCK UP, ABOUT THEM!!!!

No, we're talking as long and about what we want here. Personally, I don't use your drivers simply because they are not reentrant (local stack) and I also don't need a cache (I don't say a cache is useless!). And HimemX is better than xmgr.

I also vaguely remember that you call everyone a NAZI who tells other people what they have to do. So better be a bit careful with your words, my dear friend!

---
MS-DOS forever!

Our late comedian Rodney Dangerfield told a very funny joke: "My son kept
asking me for a bow-and-arrow for Christmas. 'Get me a bow-and-arrow for
Christmas!', and when I did, he got me a shirt with a bulls-eye on it"!

You too need such a shirt, you FOOL!! It was not Lucho who was BANNED by
Udo Kuhnt from the EDR-DOS forum. Udo banned YOUR "sorry ass", you FOOL,
because of all the BULLSH*T you tried to pull on Udo himself!! Lucho had
NOTHING to do with getting you banned --- You "blew your OWN case" on that
forum, all by your "sorry" SELF, you FOOL!!

As for Lucho "ruining the reputation of FreeDOS", their "sorry ass" system
ruined its OWN reputation, all by itself! ANYONE who doubts this can run
their OWN file copy/compare tests using FreeDOS, then again with ANY other
DOS kernel now in use! No need to take my word, or Lucho's word -- Those
with more than "half-a-brain" can TEST THE SYSTEMS THEMSELVES, so they can
SEE for themselves that FreeDOS really DOES rank "Dead LAST" in speed!!

And how FOOLISH it also is, that FreeDOS "eats" ALL available HMA space as
its disk caching buffers, leaving NOTHING for UDMA's binary-search tables,
XMS managers like MS-DOS or PTS-DOS HIMEM, and other more USEFUL programs!
Anyone who wonders about THIS "issue" is welcome to try Lucho's "LZ" V7.10
MS-DOS with "BUFFERS=1" to "BUFFERS=5" in the CONFIG.SYS file. They will
note some speed loss (about 25%) with "BUFFERS=1", a tiny speed loss (10%)
with "BUFFERS=2" and NO loss with "BUFFERS=3". To be certain, I use "4".

So Why-in-HELL does FreeDOS need to allocate 20 buffers and up?? Do they
REALLY believe that so many buffers "help"?? Especially if a CACHE is in
use, like my UDMA or even their LBACache?? Is LBACache really THAT bad??
I do not use it (obviously!), but I cannot believe even IT needs all those
buffers! Ah, but we'll never be sure, since FreeDOS has NO WAY for users
to allocate other than ALL of the HMA for buffers! I've even tried their
negative "BUFFERS = -nn" command, and it DOESN'T WORK either!

You should be more CAREFUL about disputing valid research or tests done by
others. Users who desire FACTS can test FreeDOS v.s. other systems, same
as they can make OTHER tests I have suggested on this board, and then they
will know the facts about YOU as well, you FOOL!! Then, I may not have a
need to send you a "bulls-eye" shirt -- Others may do it ANYWAY!!

> No, we're talking as long and about what we want here.

Yes, I have noticed how "SH*T Disturbers" like you DO behave that way, FOOL!
Perhaps you might succeed in getting your "sorry ass" BANNED here, as well!!

> Personally, I don't use your drivers simply because they are not reentrant
> (local stack) ...

Oh, REALLY?? Then perhaps you haven't NOTICED, FOOL, that XMGR runs on the
USER stack, same as ANY "XMS Manager" since Gates & Co. wrote the FIRST one!
Also, UDVD runs on the SHCDX33C stack as it is a "private" driver for use by
SHCDX33C only! Nor is it a problem, FOOL, for SHCDX33C (UDVD's "boss") and
UDMA to declare their own local stacks, since you obviously are unaware that
DOS was never designed for and DOES NOT SUPPORT "re-entrant" I-O, FOOL!

If you like "re-entrant" I-O, you'll LOVE "Windows", which is one reason why
Gates & Co. require "gazillion byte" drivers for it! I prefer writing 4.5K
DOS UltraDMA drivers (caching included!), especially since a friend tells me
that re: Windows, "You beat THEM, two months AGO!", and that was in July!!

> ... and I also don't need a cache (I don't say a cache is useless!).

"Oh, FRABJOUS Day!" that you found time to clarify THAT as well, for us all!

> And HimemX is better than xmgr.

Oh, REALLY?? Well, if people are so DESPERATE to save 2600 bytes of upper-
memory that they would become "tied" to using ONLY your JEMM386, they can go
ahead and keep your himemx. XMGR is not "restricted" to running ONLY with
one "EMS manager", since I believe in writing GENERAL-PURPOSE software!!

> I also vaguely remember that you call everyone a NAZI who tells other
> people what they have to do.

NOT everyone, FOOL!! Just guys like YOU, who BEHAVE like them!!

> Yes, I have noticed how "SH*T Disturbers" like you DO behave that way,
> FOOL!
> Perhaps you might succeed in getting your "sorry ass" BANNED here, as
> well!!

Hm, yes, every now and then this board's admin loves to delete all posts, might very well be that he also decides to delete all registered users for some reasons. But I'm pretty sure that he's not so "unwise" like the "Evil DoctoR" admin who usually fought all users reporting bugs (like NTOSKRNL, me ... and Lucho! IIRC). But why talking about things of the past, in the meantime dear Udo has banned the rest of the users as well.

> since you obviously are unaware that
> DOS was never designed for and DOES NOT SUPPORT "re-entrant" I-O, FOOL!

No. DOS was not designed for multitasking, but it exposes an API to make the code reusable. So a carefully written TSR might use DOS, even inside an ISR. But low-level drivers using a static local stack - like UDMA - makes this impossible. Furthermore, there are hooks/call-outs defined for int 13h drivers (int 15h, ax=9000h) to support multitasking, but UDMA doesn't implement them either.

---
MS-DOS forever!

> DOS was not designed for multitasking, but it exposes an API to make the
> code reusable. So a carefully written TSR might use DOS, even inside
> an ISR. But low-level drivers using a static local stack - like UDMA -
> makes this impossible.

Oh, REALLY?? How does UDMA do this?? Its static stack is in use ONLY
while it executes its OWN disk I-O requests, and at all other times your
"mythical TSR" damned-well BETTER have its OWN stack, FOOL!!

> Furthermore, there are hooks/call-outs defined for int 13h drivers
> (int 15h, ax=9000h) to support multitasking, but UDMA doesn't implement
> them either.

And it does not NEED to, since the RBIL clearly says that a MULTITASKER,
NOT the I-O drivers, must issue your Int 15h AH=90h "hooks"! Can't you
READ, FOOL??!! What possible "good" would those hooks be in any case??
UDMA does its OWN disk I-O and does NOT use any "Brat BIOS" logic, which
I long-ago learned NEVER to "trust"! Since the real BIOS does not know
about Int 13h requests that get intercepted by UDMA, your "hooks" are in
fact WORTHLESS, you Bloody FOOL!!

Why don't you STOP practicing that old U.S. joke: "If you can't WOW 'em
with brilliance, then BAFFLE 'em with BULLSH*T"! I will continue to do
as many posts with the TRUTH about all your BULLSH*T as necessary, until
people finally "see you" for the IDIOT and the CHARLATAN you really are!
Maybe you should follow Patrick Swayze's great words from our 1987 movie
"Dirty Dancing": STAY with your F***ing JEMM386 "College boy, and leave
the ROUGH stuff to me!"

> ANYONE who doubts this can run their OWN file copy/compare tests using
> FreeDOS, then again with ANY other DOS kernel now in use!
> No need to take my word, or Lucho's word --

I don't doubt that.

However it would be nice to include some details here like
what is average file size
what is the directory depth
what program(s) are used to copy/compare
etc.

to make comparisons more meaningful. details matter

> so they can SEE for themselves that FreeDOS really DOES rank "Dead LAST"
> in speed!!

quite possible, but the only free DOS available, so there's not much choice ;)

> And how FOOLISH it also is, that FreeDOS "eats" ALL available HMA space
> as its disk caching buffers, leaving NOTHING for UDMA's binary-search
> tables

It's intended to
allocate as much HMA space for buffers (as told by config.sys)
(and yes, in case a cache driver is present buffers=5 is probably fine)
allocate the rest for buffers as well, until someone else asks for it (that's the way I designed it)

BUT: there is indeed a problem here.
AFAIR this doesn't work during config.sys processing, as Freedos goes to HMA, initializes buffers etc. only after all device='s have been processed.

(re-)using HMA buffers should be fine when UDMA is loaded by devload UDMA.SYS (or by running a hypothetical UDMA.EXE)


B) I tested UDMA/UDVD 9-23-07 on a NFORCE4 motherboard, with S-ATA disk, and S-ATA DVD; both work 'as advertised'
disk with ~60 MB/sec
DVD with ~11 MB/sec

thanks, great work again

c) 2 minor 'buglets':
if UDMA is loaded twice (may happen if playing with config files), it just loads a second time. maybe better to just abort with 'already loaded'

AFAIK all XMS managers register a device name 'XMSXXXX0'

with XMGR, the following test fails
if exist XMSXXXX0 goto _xmsok
echo.
echo AUTOEXEC: No XMS manager installed, can't init RAMDISK
goto _abort

last and least, when UDMA and UDVD are loaded, the list of IDE adapters
is shown twice.
would be nice if the latter could be suppressed (reduce verbosity)

nothing of that a big deal, of course.

Anyway, thanks for the drivers, and thanks for adding S-ATA support.

tom

> It would be nice to include some details here, like what is average file
> size, ...

Ask those who still use FreeDOS. I do not.

> It's intended to allocate as much HMA space for buffers ... BUT there is
> indeed a problem here.

Glad to hear you concur.

> Using HMA buffers should be fine when UDMA is loaded by "devload
> UDMA.SYS" (or by running a hypothetical UDMA.EXE)

Just FIX your problem. Few use "devload", and UDMA.EXE is otherwise not
needed, as it has NO "run-time" user options.

> I tested UDMA/UDVD 9-23-07 on an NFORCE4 motherboard, with S-ATA disk,
> and S-ATA DVD; both work 'as advertised', disk with ~60 MB/sec, DVD with
> ~11 MB/sec. Thanks, great work again.

A Bloody MIRACLE, seeing such words by YOU after "past events" that I hope
I need NOT remind you about! Tell "friend" Grech a good SATA DVD test is
now completed, as he seemed to want to hear.

> if UDMA is loaded twice (may happen if playing with config files), it just
> loads a second time. maybe better to just abort with 'already loaded'

Costs init code, and I have an obligation to Lucho's full "boot" diskettes
that UDMA does not go over 4.5K in size. The first copy of UDMA shall be
ignored after the 2nd "hooks" Int 13h. Tell your people "Don't DO this!"

> AFAIK all XMS managers register a device name 'XMSXXXX0'

That feature was deleted, when I eliminated all "baggage" and went back to
"personal" drivers, after OTHER problems with your countrymen! NEVER was
my intent to have public drivers again, until I simply felt sorry for many
"Good Guys" who use DOS. Now, you can test for XMGR2$ or XMGRB2$ (boot).

> last and least, when UDMA and UDVD are loaded, the list of IDE adapters
> is shown twice.

Some people do run UDVD without UDMA, and UDVD must display its own list.
UDVD must also save init logic, since sooner-or-later the drivers will be
REQUIRED to deal with ADMA/AHCI, like they had to begin running SATA now!
Go "have some coffee" while the second controller list is displayed.

> Ask those who still use FreeDOS. I do not.
>
> Just FIX your problem.
>
> A Bloody MIRACLE, seeing such words by YOU after "past events" that I
> hope I need NOT remind you about! Tell "friend" Grech a good SATA DVD test
> is now completed, as he seemed to want to hear.
>
> Tell your people "Don't DO this!"
>
> I simply felt sorry for many "Good Guys" who use DOS.
>
> Go "have some coffee" while the second controller list is displayed.

You may be the bloody genius you say you are, but you are a very bad colleague.

I don't come her looking for "truth", but for facts and for the informed opinions of other DOS users. There's room even in technical discussion for human feeling, joking, a bit of satire perhaps, but you go far beyond any of that. You are perfect, anyone who has a problem with any of your programs is a fool or a corrupt foreigner. As far as I'm concerned, you can "hit the road, Jack, and don't you come back no more, no more, no more, no more". (How's that for a good American quotation?)

First, why don't you read Tom Ehlert's 10-Feb-2006 FD-Devel post, where
he calls XCDROM "stupid", after having praised UDMA (he still has it in
his software distro) and without even ASKING me beforehand if there may
have been REASONS for why XCDROM differed from XDMA.

From his post in 2006, I had REASONS today for still feeling just-a-bit
"skeptical" about Tom, which is why I replied in such a "tone". Also,
I refused Tom's requests because there ARE "bigger fish to fry". UDMA
cannot go over 4.5K because Lucho's "boot" diskettes are LOADED, with 0
spare sectors even AFTER using the 2 diagnostic cylinders that few even
KNOW about! And UDMA still must "deal with" ADMA/AHCI controllers for
SATA, with 81 bytes left! I will find a way, but it will NOT be easy!

I do NOT use FreeDOS. I and many others have tried it in the past and
it really IS terrible, BIG time! What is wrong with saying THAT??

I really DID feel sorry, that other DOS users did NOT have a cache like
UDMA. It has taken four years to go from a barely-working single disk
driver to a full-cache driver that handles 250-MB easily and 1-Gigabyte
maximum! I am PROUD of UDMA and its companions. They DID take much
HARD WORK. They are absolutely NOT any sort of "genius" creation!

Now, you show up.

And you think saying all the above means I believe myself "perfect".

And now you say I should follow Ray Charles' words and "Hit the road"!

Fine, Steve. Take note of what is now on Johnson Lam's website.

The latest 28-Sep-2007 drivers are withdrawn. The prior drivers will
be retained, to help people NOT as unreasonable as some on this board.

Excepting my 3 good advisors and Johnson's kind friends in China, I do
NOT believe I will find myself "feeling sorry" for DOS users any more.

And if you wonder why all this happened, Steve ...

> And if you wonder why all this happened, Steve ...

Because you are a very unpleasant person.

> Software cannot be stolen. It can just be copied, unlike physical
> objects.

Utterly meaningless.
1) A copyrighted work is property under law.
2) Appropriating a work in violation of the copyright holder's conditions is stealing.
3) Physical objects can be copied. Ask any child.

> only a court can say what's legal and what's not.

When you put it that way, no.

> > Furthermore, there are hooks/call-outs defined for int 13h drivers
> > (int 15h, ax=9000h) to support multitasking, but UDMA doesn't implement
> > them either.
>
> And it does not NEED to, since the RBIL clearly says that a MULTITASKER,
> NOT the I-O drivers, must issue your Int 15h AH=90h "hooks"! Can't you
> READ, FOOL??!!

You don't know what you're talking about, my friend. The Int 13h driver is supposed to issue int 15h, ax=9000h, which will indicate for a multitasker that it has to wait for the I/O operation to complete. Consequently, the (irq 14/15) ISR is supposed to issue int 15h, ax=9100h to finally indicate that the io operation is finished.

The MULTITASKER is supposed to HOOK into the int 15h chain, but the Int 13h driver must CALL it. Is that so hard to understand? Any other behavior simply doesn't make any sense.

Well, since I'm a gentle person I'm hereby making you a offer to teach you the basics of the ROM-BIOS. Will cost you nothing, at least the first 2 lessons.

---
MS-DOS forever!

> Well, one example are your little attempts to ruin FreeDOS's reputation, my big hero Lucho
> (http://www.drdosprojects.de/forum/drp_forum/posts/4363.html).

I am by no means neither yours nor a hero, and I'm not sorry for the above post.
If saying the truth is ruinning reputation, your attitude to FreeDOS is rather strange. (There's a better four-letter word for this, but I won't say it here.)

> A small hint, my friend: unlike the "Evil DoctoR" board you won't be able
> to hide behind the back of the admin here.

Never did that, and no need for me to do it.
Don't be double-faced to call me a "friend".

You'd better stop your flame war against me and Jack and don't be a "troll". You won't gain anything, even with Steve's help. Quite the contrary, you discredit yourself, as you've done on the Udo's forum.

Accept this friendly advice from a non-friend who is older than you (I'm 48).

Tom's post here is a good example of constructivism. Try to follow it if you can

> I don't come her looking for "truth", but for facts and for the informed
> opinions of other DOS users. There's room even in technical discussion for
> human feeling, joking, a bit of satire perhaps, but you go far beyond any
> of that.

That's absolutely right.

> You are perfect, anyone who has a problem with any of your
> programs is a fool or a corrupt foreigner. As far as I'm concerned, you
> can "hit the road, Jack, and don't you come back no more, no more, no
> more, no more". (How's that for a good American quotation?)

Now you step over exactly the same lines that you critize above.

I don't know who you are and where you came from.
But it would kind if you would leave this forum into the direction
where you came from.

> I am by no means neither yours nor a hero, and I'm not sorry for the above
> post.

This was meant ironically. Just to make it absolutely clear, even for the simple-minded.

> If saying the truth is ruinning reputation,

It doesn't really matter if it is true or not. The important thing was the intention behind your words, and this intention is easy to see, it is "destructive".

> You'd better stop your flame war against me and Jack and don't be a
> "troll".

Jack is innocent, but you are a thief and a self-rightous coward. Hopefully this is clear enough. Indeed it was you who started all this bullshit once in the "Evil DoctoR" forum. A pest.

> Tom's post here is a good example of constructivism.

Sure, but a waste of time. He also has "crossed the Styx", that is, he's one of the "bad guys" and any positive comment on the drivers is just "damage control" in Mr. Ellis eyes.

---
MS-DOS forever!

> Utterly meaningless.
> 1) A copyrighted work is property under law.
> 2) Appropriating a work in violation of the copyright holder's conditions
> is stealing.

Instead of trying to dispute this with you, let's see what the FSF has to say:

----- start of quote -----

?Intellectual property?

Publishers and lawyers like to describe copyright as ?intellectual property??a term that also includes patents, trademarks, and other more obscure areas of law. These laws have so little in common, and differ so much, that it is ill-advised to generalize about them. It is best to talk specifically about ?copyright,? or about ?patents,? or about ?trademarks.?

The term ?intellectual property? carries a hidden assumption?that the way to think about all these disparate issues is based on an analogy with physical objects, and our ideas of physical property.

When it comes to copying, this analogy disregards the crucial difference between material objects and information: information can be copied and shared almost effortlessly, while material objects can't be.

To avoid the bias and confusion of this term, it is best to make a firm decision not to speak or even think in terms of ?intellectual property?.

The hypocrisy of calling these powers ?rights? is starting to make WIPO embarassed.

?Theft?

Copyright apologists often use words like ?stolen? and ?theft? to describe copyright infringement. At the same time, they ask us to treat the legal system as an authority on ethics: if copying is forbidden, it must be wrong.

So it is pertinent to mention that the legal system?at least in the US?rejects the idea that copyright infringement is ?theft.? Copyright apologists are making an appeal to authority ? and misrepresenting what authority says.

The idea that laws decide what is right or wrong is mistaken in general. Laws are, at their best, an attempt to achieve justice; to say that laws define justice or ethical conduct is turning things upside down.

----- end of quotes -----

The above quote is from http://www.gnu.org/philosophy/words-to-avoid.html

> > If saying the truth is ruinning reputation,
>
> It doesn't really matter if it is true or not.

Good "logic"! And that was written by a programmer?! What a shame!

> Jack is innocent, but you are a thief and a self-rightous coward.

Very bold from you to insult blatantly, knowing that you won't be sued for that!

> Indeed it was you who started all this bullshit once in the "Evil DoctoR" forum. A pest.

So you finally fell to the lowest possible level - insulting personal qualifications. Very well. Thus you fully discredited yourself, once again.

You can continue with your personal attacks, if this is so pleasant to you... Come on, spit more venom! All your venom (even risking to drown yourself in it)!

DR-DOS is not evil. You are!!!

> Very bold from you to insult blatantly, knowing that you won't be sued for
> that!

Since you apparently don't understand irony and sarcasm ...

Well, suing me ... would possibly be a not so good idea as long as you still distribute your MS-DOS clone (wer im Glashaus sitzt ...).

But ok, I've told what I had to say. Continue with your work, you "good guy"!

---
MS-DOS forever!

> Now you step over exactly the same lines that you critize above.
> I don't know who you are and where you came from. But it would
> be kind if you would leave this forum into the direction where
> you came from.

No need, Tom. I will leave.

And you know the sad thing in all this. After having replied to
you as I did, the very NEXT time I approached my computer, I WAS
ready to admit that if you WERE sincere this time, I WOULD have
worked with you. You know disk drives from the hardware side, I
know disk drives from the software side, and we might have made
both your "Snapshot" systems and my drivers a lot better.

And on logging-in 12 hours ago, the FIRST thing I see is THIS guy.

Johnson, Lucho, and my other advisor (wishes to remain anonymous)
have now been told to handle my drivers as "private" again. The
latest update, wherein UDVD doesn't require a /C switch any more,
is now "canned" from Johnson's server, as you can see. And I've
also written all 3 of those men that they will NEVER need to tell
me again "You cannot win!", dealing with boards like this.

Sorry, Tom. I will leave.

> Instead of trying to dispute this with you, let's see what the FSF has to
> say:
>
> ----- start of quote -----
>
> ?Intellectual property?
>
> Publishers and lawyers like to describe copyright as ?intellectual
> property?

blah blah blah

That's not a legal analysis or an accurate summary of copyright law. It's a subjective moral objection to copyright laws, and doesn't change the fact that some of us live in countries that have those laws.

> You'd better stop your flame war against me and Jack and don't be a
> "troll". You won't gain anything, even with Steve's help.

I am not a flamer (at least not here, not yet), a troll, or Japheth's helper (that he and I appear to have similar attitudes toward the real flamers is a fact but not a conspiracy).

> Accept this friendly advice from a non-friend who is older than you (I'm
> 48).

Well, I'm older than you. So are you going to obey me?

> Tom's post here is a good example of constructivism. Try to follow it if
> you can.

Shouldn't that be addressed to Jacko?

> > You are perfect, anyone who has a problem with any of your
> > programs is a fool or a corrupt foreigner. As far as I'm concerned, you
> > can "hit the road, Jack, and don't you come back no more, no more, no
> > more, no more". (How's that for a good American quotation?)
>
> Now you step over exactly the same lines that you critize above.

I didn't step over any line. Maybe you haven't read all of his messages.

> I don't know who you are and where you came from.
> But it would kind if you would leave this forum into the direction
> where you came from.

Now you're doing it. You didn't think so, did you?

See what I mean?