30035 9-JUN 14:40  General Information
     RE: Pictures Needed (Re: Msg 29980)
     From: DICKWHITE    To: DALEP (NR)

Well, I will start doing some digging.  Hope I don't come across any
undecomposed hot dog or partially empty cat food cans.

-*-

30036 9-JUN 15:07  Graphics & Music
     RE: framed window within a framed window (Re: Msg 30008)
     From: EDDIEKUNS    To: ZACKSESSIONS

I have only one thing to add.  (Some general comments about get/put buffers. I
know YOU know this, Zack, but ...!)

Get your PID, and use that as your group number.  Before you do any block
manipulation, Kill bufer 0 in your group to empty the entire group.  Just in
case.  When you are done, kill it again.

                           Eddie

-*-

30040 9-JUN 16:47  Graphics & Music
     RE: framed window within a framed window (Re: Msg 30008)
     From: NES          To: ZACKSESSIONS

Humm...  I think I'll use the new cgfx command call MVFName(), as soon as I can
figgure out how to call it... -nes

-*-

30045 9-JUN 19:08  Graphics & Music
     RE: framed window within a framed window (Re: Msg 30040)
     From: DODGECOLT    To: NES

Well, I wrote the CGFX lib in the databases here. The newest version has fairly
complete docs. What do you need to know about the function?
 -Mike

-*-

30050 9-JUN 22:38  Graphics & Music
     RE: framed window within a framed window (Re: Msg 30040)
     From: ZACKSESSIONS To: NES

Just remember that that call is "non-standard" and a custom function designed by

Mike Sweet for his "replacement" cgfx lib.

Zack

-*-

30088 10-JUN 10:17 Graphics & Music
     RE: framed window within a framed window (Re: Msg 30045)
     From: NES          To: DODGECOLT

Mike, I think I have your newest version of the cgfx.l, with file selection
rutines, My queston is now do you call MVFName?

MVFName(stdout,"FILES",2,1)? or ptrname=MVFName(STDOUT,"FILES",2,1); I am
getting a Stack OverFlow when I call it.  I am still new to C and have trouble
with pointers. -Eric


-*-

30094 10-JUN 17:30 Graphics & Music
     RE: framed window within a framed window (Re: Msg 30088)
     From: DODGECOLT    To: NES (NR)

Ok, here is an example of how you should call MVFName()-

 main()
 {
  char *name;

  SetGC(1,GRP_PTR,PTR_ARR);
  _ss_mous(1,3,10,1); /* sample every 3 clock ticks, 10 tick timeout, follow */

  name=MVFName(STDOUT,"Filename?",10,4,0,2);


  if (name==NULL) /* no filename selected...*/
   exit(216);  /* exit here... */


  /* rest of code follows (say, to list the selected file... */

 _ss_mous(1,0,0,0); /* turn off auto-follow */
 SetGC(1,0,0); /* turn off mouse pointer */
 }

 I hope this will help clear things up.  I guess I really ought to include
examples in the docs for the file-picking functions also.
 -Mike

-*-

End of Thread.

-*-

30037 9-JUN 15:08  Programmers Den
     RE: Aciapak bugs.  Chapter 3.  <Sigh> (Re: Msg 30013)
     From: EDDIEKUNS    To: MAREK

Yes, actually, it WAS interesting!

                Eddie

-*-

30039 9-JUN 15:17  Programmers Den
     RE: Aciapak bugs.  Chapter 3.  <Sigh> (Re: Msg 30028)
     From: EDDIEKUNS    To: KNOT1

Yes, I've tied all the IRQ lines in my Multipak together.  It's possible that I
have a bad solder joint tho.  I'm gonna think on the rest of the stuff you said
(thanks!) and see if I get any brainstorms

                     Eddie

-*-

30077 10-JUN 03:04 Programmers Den
     RE: Aciapak bugs.  Chapter 3.  <Sigh> (Re: Msg 30028)
     From: OS9UGPRES    To: KNOT1 (NR)

Jamie,

Yeah, I use that "locked up" trick a lot to send select (1b21) to a window I
can't see. "echo 1b21 >/wx" and then hit BREAK a coupla times. Up comes the
hidden screen, and then I hit ENTER on it to keep it alive.

Doesn't make sense to me either. The first part, anyway. The last part does.
Maybe someone can help figure this out. One of the factors is that SCF drivers
send signals (Break) to the LAST process to use a device...so:

 echo hello >/w2   - Locked up (echo running on /w1, let's say)
                     Note that Echo has done no I/O to /w1.

 BREAK (w1)        - Last output process was the shell on /w1,
                     so that shell sends Kill signal to "echo".
                     Echo was waiting on /w2, tho. Can't be killed yet.

 BREAK (w1)        - Somehow causes echo to get thru. Dunno how!

 BREAK (w2)        - Now last process to use w2 was Echo! So it gets
                     this Break, and you see "Error 002" on /w1.

I'll have to ask Kent Meyers why Echo can ever get thru. I can only guess that
SCF is letting Echo get the output device (/w2) while the shell on that window
has the INPUT side of /w2 locked up. SCF does some weird stuff trying to get
both the input/output sides of a device allocated. - kev

-*-

End of Thread.

-*-

30038 9-JUN 15:13  General Information
     RE: New C compiler bug!  (newly found) (Re: Msg 30020)
     From: EDDIEKUNS    To: RAGTIMER

Yeah -- it's frightening how much power the Intel gremlins hold, isn't it. I
can't imagine not being able to write position independent, non-reentrant code!
(Which I hear the 386 can finally do?  I know very little about the Intel chips
at the actual instruction level.)

                      Eddie

-*-

30046 9-JUN 19:40  General Information
     RE: New C compiler bug!  (newly found) (Re: Msg 30021)
     From: GREGL        To: RAGTIMER

Mike,

    That is the biggest pain in the derrier. Not only does it cause headaches
with graphics bit-mapping, it also causes major headaches in other areas.
Although a graphics bit-map isn't that much of a problem; mostly swap the
attribute/character values for text mode displays and the actual bit-map is a
continuous stream of binary data. That is until you hit the bit-planes used in
EGA and VGA.
    I almost had AR working except the reverse integer ordering really threw off

his LZW algorithms. One day I might attack that with a routine to convert
big-endian to little-endian on the fly but it will be a royal pain to do;
especially consider the LZW routines used. But I think the biggest blunder is
the fact that is has "Intel" on the chip.

    -- Greg

-*-

30047 9-JUN 20:06  General Information
     RE: New C compiler bug!  (newly found) (Re: Msg 30038)
     From: GREGL        To: EDDIEKUNS

Eddie,

    I think position-independent code on the 8086-series is most a fallacy. On
the 8086, 80186 and 80286 there isn't much, if anything, that uses offsets
instead of absolute addressing. You can get around that by adjusting the values
in the segment registers. Each "segment" occupies 16-bytes of memory; that is,
segment address 0000:0000 is physical address $00000 and segment address
0001:0000 is physical address $00010. The largest address the 8086 can access is

$FFFFF, which is a segment address of F000:FFFF and/or FFFF:000F. You can obtain

the physical address by shifting the segment register left 8-bits and adding the

offset.
    So, while each instruction refers to code or data with a physical address it

is really an offset within the segment. The 80286 still uses the segmented
architecture even in protected mode as does the 80386. Although the 80386 has
allows you to disable the segmentation to use a virtual linear address space.
But as you are probably well aware, the segmented architecture of the
8086-series gets in the way of operating sytems and is probably the best reason
a decent operating system has never been written for it. It's a real mess, no
lie.
    Another problem with the 8086-series is that the time needed to perform
branches in protected mode really jumps sky-high. A JMP instruction normally
takes 11 clock cycles; it takes 23 clock cycles in protected mode to perform a
JMP within your own code. A JMP instruction to a "Call Gate" (or "Task Gate")
takes 183 clock cycles. A Task Gate is used to branch from one module to
another. A software interrupt via a Task Gate (such as a system call) takes 167
clock cylces in protected mode. Those are on the 80286, by the way. The 80386
really cuts down the time required by buffering the most-used call gates and
such on-chip.

    -- Greg

-*-

30053 10-JUN 00:45 General Information
     RE: New C compiler bug!  (newly found) (Re: Msg 30047)
     From: RAGTIMER     To: GREGL (NR)

Task gate?  Call gate?  Sounds like the movie 2001 ending to me! Or an even
worse kludge than the subroutine modules of OS9 that I use to fit Ultimuse-3
into the 64K address space.

Actually, all the OS9 ballyhoo about position-indpenedent code can be eliminated

just by having relocation registers, which any multi-taksing CPU ought to have.
Ya know, with the GIME, you could just as well compile absolute addreses into
tasks (but you couldn't merge a bunch of them into one RAM block, so...).

I hope the 68020 or 030 can do relative jumps bigger than 32 K, by the way.
Wait till you get OSK and read out the linker kludges ABSOLUTE jump addresses
into the code.  It work, it ain't to slow, but it sure ain't OS9 like we learned

it, grin.

-*-

30058 10-JUN 01:32 General Information
     RE: New C compiler bug!  (newly found) (Re: Msg 30047)
     From: EDDIEKUNS    To: GREGL (NR)

Does this mean that programs which are system call intensive will actually run
faster on a CoCo than on a '286?  <Grin>

How long does an OS-9 system call take to process?  (Not the actual system call
... the switch from running your code to running the system call code.)

Those are frightfully large numbers, BTW!!!  It makes the nine cycles you wait
for an LBSR seem pretty trivial!

                  Eddie

-*-

End of Thread.

-*-

30041 9-JUN 16:50  General Information
     RE: MM1 demo (Re: Msg 30031)
     From: NES          To: SANDRIDER

The gif's where display veary quikly, in 1 sec or faster depeneding on the
display mode. I didnt realy time it so just a guess.. -nes

-*-

30081 10-JUN 04:11 General Information
     RE: MM1 demo (Re: Msg 30041)
     From: SANDRIDER    To: NES


  One second, huh?  Good enough for me!

            -CRW

(sure beats three minutes!  To be fair, though, the CoCo viewer does have to
 do a lot more thinking about it.)

-*-

End of Thread.

-*-

30042 9-JUN 17:12  Programmers Den
     RE: Shell+ 2.1 and Vi (Re: Msg 30014)
     From: RICKADAMS    To: MAREK

Interesting.  I use vi with Shell+, but it's an older version, I think.

-*-

30049 9-JUN 21:04  Programmers Den
     RE: Shell+ 2.1 and Vi (Re: Msg 30018)
     From: MAREK        To: ZACKSESSIONS

oops..got a little confused last night...NO I do NOT load VI it is included in
my OS9Boot file and with AGIVIRQDr.  but since VI is a device descriptor i
assumed that OS9 would know that the memory resident VI is a dd and not an
executable file.

I hope that clears that up.

I'll os9gen a new boot disk without VI and the associated driver and let you
know what happens.

-Michael


-*-

30055 10-JUN 00:53 Programmers Den
     RE: Shell+ 2.1 and Vi (Re: Msg 30049)
     From: RICKADAMS    To: MAREK

OOPS.  Yes, if the Kings Quest "vi" module is in your boot, you won't be able to

run the "vi" program.  The system doesn't know one from the other.

-*-

30057 10-JUN 01:32 Programmers Den
     RE: Shell+ 2.1 and Vi (Re: Msg 30055)
     From: MAREK        To: RICKADAMS

YEP!  Your answer was correct...removed VI and TSEdit/vi fired up...one problem
though,  I was editing a file and printed it out with the :p option...not all of

the text printed out!  I checked the file and it was all there and tried to
print it again with the :p option same problem but this time it was missing
different lines.  What could be causing this?

-Michael


-*-

30092 10-JUN 15:46 Programmers Den
     RE: Shell+ 2.1 and Vi (Re: Msg 30057)
     From: RICKADAMS    To: MAREK (NR)

I have had the same problem with TSEDIT not printing the entire file with the :p

command.  I have NO idea what the problem is, and I had it before I came up with

the patches, too.

I've just taken to exiting TSEDIT and using "list filename >/p" to print the
file.

-*-

End of Thread.

-*-

30043 9-JUN 17:54  General Information
     RE:  Stuff for Sale (Re: Msg 29953)
     From: MPASSER      To: GENEDEAL (NR)

Gene,

Yes, I would love to have it... I will leave you my address/phone/etc. in MAIL.

Thanks!

Mike Passeer

-*-

30044 9-JUN 18:44  General Information
     RS Hard Drive
     From: TEDJAEGER    To: GREGL

Greg, I have been trying to get a RS HD interface to work in a replacement
product for my old (large) MPI interface. Most recently I have simply tried
replacing the old MPI with a newer (small) one. Much to my surprise, while
everything works fine under os9 level II using my old MPI, when I install the
new one, the hard drive will not boot--I get the nice green screen with "BOOT
FAILED." I have tried two small MPIs now, one not upgraded for CoCo3 that I got
upgraded and one that was upgraded to start with. If I boot into a floppy system

all is fine but when I have renamed the hard drive descriptor hd and tried to
access the hard drive through /hd I have consistently received error 240 -- unit

error. I am using a CoCo 3 and have tried my hardware on another CoCo 3 at the
local Radio Shack store with the same results. It really seems as if Radio
Shack, when they redesigned the MPI, did something to make it itcompatible with
their own HD interface. I seem to remember that you use(d) one of those 35 meg
RS hard drives. I was wondering if you used a small or large MPI and if you had
any ideas why a newer MPI should fail where an older one that was upgraded would

work. --Thanks, TedJaeger

-*-

30048 9-JUN 20:10  General Information
     RE: RS Hard Drive (Re: Msg 30044)
     From: GREGL        To: TEDJAEGER

Ted,

    I use one of the older and larger Multi-Paks with the 35-meg hard drive. I
haven't tried using one of the newer and smaller models. Come to think of it, I
don't think we have one of the smaller models. You might ask Marty, I'm not
familiar with the new Multi-Pak at all.

    -- Greg

-*-

End of Thread.

-*-

30051 9-JUN 23:15  Telcom
     RE: OSTERM (Re: Msg 30023)
     From: ELM          To: RICKADAMS

Ok, I'll try logging on with the space . space.

Thanks!

-Len

-*-

30059 10-JUN 01:34 Telcom
     RE: OSTERM (Re: Msg 30051)
     From: EDDIEKUNS    To: ELM

I've noticed that Delphi acts DIFFERENT if you login with CR . CR!  You don't
get that TERMINAL= prompt!  And when you type ^Z you get the control Z echoed
before Delphi echos the ^ and Z characters.  Odd.

                     Eddie

P.S.  Or should I say TELENET acts different?  :-)

-*-

30087 10-JUN 10:10 Telcom
     RE: OSTERM (Re: Msg 30059)
     From: ELM          To: EDDIEKUNS (NR)

It's interesting that since I posted my question about logging on with 8N1 I
have gotten several answers, all different.  I haven't tried them all yet (I did

try one that didn't work:  responding to the onscreen garbage with "a").

I would think that Telenet is likely the problem, and I seem to recall there
being a Telenet helpfile.  I wonder if anyone knows how to access the Telenet
help?

I've also been told that different Telenet nodes have varying characteristics
and that what works on one won't work on another.

Truly a miracle of modern telecommunication.  Makes one think seriously about
going back to smoke signals (7 baud).

Thanks for the help.

-Len

-*-

End of Thread.

-*-

30052 10-JUN 00:38 Programmers Den
     RE: RMA and Debug (Re: Msg 29890)
     From: RAGTIMER     To: EDDIEKUNS

Yeah, get a book.  There are 8 Data regs and 8 Address regs, all 32 bits (Wow!
as Drew A Kaplan would say).  Forget the funny 8/16 bit stuff (tho it was a nice

way to break bytes out of words.

-*-

30072 10-JUN 03:03 Programmers Den
     RE: RMA and Debug (Re: Msg 29890)
     From: OS9UGPRES    To: EDDIEKUNS (NR)

The 680x0 series has 8 x 32-bit data regs, and 7 x 32-bit address regs (the 8th
address reg is the Stack Pointer).

The data regs are called "d0-d7", and the address "a0-a7". The condition codes
reg is called "ccr".

The data regs can hold byte (8-bit), word (16-bit) or long word (32-bit) values,

and can usually do ops on all three types. Those three modes are appended to the

end of the op you want to do. For example:

 add.b #2,d0  - add $02       to d0's lowest 8-bits (byte)
 add.w #2,d0  - add $0002     to d0's lowest 16-bits (word)
 add.l #2,d0  - add $00000002 to d0   all 32-bits (long word)

The default, if you don't specify, is .w for word. However, most people always
add the .suffix just so they can more easily check things out. I've already been

hit by not doing it... I had left it off when I really wanted to do a .b
operation... and had weird crashing problems as a result of my goof. It wasn't
readily apparent without the suffix.

-*-

End of Thread.

-*-

30054 10-JUN 00:48 Graphics & Music
     RE: MVCanvas 2.0 (Re: Msg 30025)
     From: RAGTIMER     To: MIKEHAALAND

OK, now I read your Forum mail, AFTER readingt the private Email. I'll go and
see if you uploaded it agian, sounds like you did. Thanks, mike k PS:  Good, I
need an excuse to unstick the ink jets in the 220 again.

-*-

30071 10-JUN 03:02 Graphics & Music
     RE: MVCanvas 2.0 (Re: Msg 30054)
     From: OS9UGPRES    To: RAGTIMER (NR)

Mike - what and how do you unstick your 220 ink jets? Mine has been clogged up
for literally years!

-*-

End of Thread.

-*-

30056 10-JUN 00:53 General Information
     RE: Pictures Needed (Re: Msg 30032)
     From: DWHILL       To: DALEP (NR)

I'm glad to see KISSable OS-9 will be back, too.  That's usually the first thing

I go looking for when I get my newest issue of Rainbow.

--Damon

-*-

30060 10-JUN 02:09 General Information
     RE: Shell+ (Re: Msg 30026)
     From: BRIANWHITE   To: KNOT1 (NR)

Good point...  Hadn't thought of that.

                                       Brian

-*-

30061 10-JUN 02:13 General Information
     For Sale
     From: BRIANWHITE   To: MPASSER

Mike,

How does $5 +S&H grab ya?  And if you think that's a little steep, I might just
settle for the S&H.

                                                           Brian

-*-

30062 10-JUN 02:13 General Information
     Shell+
     From: BRIANWHITE   To: ZACKSESSIONS

Zack,

Yea, I heard something about an ALIAS command.  I jsust assumed it created a
tiny batch file that forked the program you asked for (assuming the shell
supported parameter passing to scripts).  I guess as long as an "EX" sits before

the command to call, n o extra space would be required by the new shell.  Of
course, then you couldn't alias one name to multiple commands.  I just thought
it would be nicer, more versitile, and faster, if the shell did it on its own.

                                                           Brian

-*-

30091 10-JUN 14:56 General Information
     RE: Shell+ (Re: Msg 30062)
     From: ZACKSESSIONS To: BRIANWHITE (NR)

Actually, it creates a tiny binary program which forks the process for you.

-*-

End of Thread.

-*-

30063 10-JUN 02:14 General Information
     Shell+
     From: BRIANWHITE   To: EDDIEKUNS (NR)

Eddie,

I agree whole-heartedly.  I've used VMS and, though I can't say I look forward
to every day I spend with it, it did have its good points... and the retrieve
was one of them.  I also have that PD program at work and it is really nice.
The only reason I d idn't suggest it is because that kind of response would have

to be build into SCFMan (Kev?).  There is no way Shell could do that without
skipping the OS-9 ReadLn command completely and just doing single char reads and

checking for up/down arrows while e mulating the OS-9 ReadLn.

All that stuff for sale is stuff that just sits in my closet.  I was hoping to
make a bit of $$$ to put towards the new computer.  The only decision left is
what computer to get.  I'd like a TC9/MM1, but I really like the high qualility
software availabl e for IBM /386.  Tough to be caught between what you need and
what you love.  I'll probably get the TC9 and then try and write a
Norton-Utilities type thing...  We'll see.

                                                           Brian

P.S.  If you really want to see versitile addressing modes, pick up a 68020 (+)
book!!!  I lent my 030 book to a friend, but I think one of the biggies is
something like: SCALED REGISTER INDEX INDIRECT WITH BASE ADDRESS AND OFFSET

A "MOVE" instruction between two address modes like that is 11 WORDS long!!!

-*-

30064 10-JUN 02:14 General Information
     RO-I
     From: BRIANWHITE   To: TJMARTIN (NR)

Sorry to say, but Robot Odyessy can't run unde L2.  The author did some pretty
sneaking tricks like passing pointers in system direct page ram.  Therefore,
because under L2 you don't have access to L2 system RAM, the paramters don't get

passed and everyt hing crashes.  However, I did successfully get ROI to run at
2MHz with the IOMan patches for the original OS-9 L1.  Big improvement there.

As for me, I'm sitting on level-4 with the mine-room and map-room.  Something
went drastically wrong with my game and I have not been able to locate another
copy.  The inovation lab stopped working to the point that everytime I burn a
chip, really wierd thigs start to happen.  I also have never been able to read
the chips that came with the game.  I had to re-design them all from scratch.
Now I don't play it because of the innovation lab problems.  It's really a wierd

problem.  All of the CRC checks ar e fine, so I'd assume all the code is
correct.  Maybe you could send me the CRC's for your version so I can check them

against mine?  Sorry I can't help with that last room...

                                                           Brian

-*-

30065 10-JUN 02:15 General Information
     RMA
     From: BRIANWHITE   To: THEFERRET (NR)

Philip,

RMA works just fine with predecriment and postincrement.  I use it all the time.

One problem you might be having is that there is no such beast as ",x-" or
",x--" on the 6809, just as there are no ",+x" or ",++x".  Just be careful when
saying they are t he same, because they are not.  When the sign is before the
register, the inc/dec comes before the translation to an effective address
instead of after as when the sign is after the register.  Just because Edtasm
interprets them the same does not mean th at they ARE the same.  It's just that
the preincrement and postdecriment aren't supported on the 6809.  Not much of a
problem, though.  I've been writing in assembly for 6 years and never found a
place where these two modes were better than the others.

                                                           Brian

-*-

30066 10-JUN 02:15 General Information
     RMA
     From: BRIANWHITE   To: TIMKOONCE (NR)

Tim,

One correction to your comment about the single increment and decriment with LDD

and STD, etc.  They do work... I use them all the time for shifting.

sftloop  ldd   ,x+
         REPT  ?
         lslb
         rola
         ENDR
         sta   ,y+
         dec   <counter
         bne   sftloop

As far as I know, this is the most efficient shift algotithm available on the
6809, or any Big-Endian chip.  (And those Intel guys just can't do it!  :)

                                                           Brian

-*-

30067 10-JUN 02:15 General Information
     OS-k RBF
     From: BRIANWHITE   To: OS9UGPRES

Kev,

Is OS-k versitile enough to handle RBF devices with different sector sizes?  I
heard OS-9 would only do it with patches to IOMan for the data size it gave to
RBF.  Or was it RBF for the size it gave to HDisk?  Either way, I have a 100Meg
hard drive that only supports a minumum of 512 byte sectors that I'd really like

to use.  I'd also suspect that the CD-Rom sector size is 2048 bytes and I'd like

to use one of those babies, too.  As far as I know, the change in sector size
can be accomplished thru the d river and specifying a cluster size of 2/8 (for
the above sizes, respectively), but I have yet to test this theory.  How close
am I?

                                                           Brian

-*-

30070 10-JUN 03:00 General Information
     RE: OS-k RBF (Re: Msg 30067)
     From: OS9UGPRES    To: BRIANWHITE (NR)

Brian - the very latest RBF knows about different sector sizes, yes. The size is

specified in the device descriptor.

That actually almost caused me to screw up the HD on my MM/1 prototype, btw. I
was mucking around and decided to use a disk editor on /H0.

What I did not realize was that /H0 was using 512 byte sectors! And guess what
size the disk editor was set up for? Uh-huh, 256 byte sectors. So I changed
something and wrote it out... but things got flaky for a while. Got me why.

-*-

End of Thread.

-*-

30068 10-JUN 02:16 General Information
     CoCo-4/OS-9
     From: BRIANWHITE   To: PAULSENIURA (NR)

Paul,

   Ditto!


There were some points you were a little off about, one of which I'd like to
comment about is the MM1 & TC9...  While I haven't decided which I will get (if
indeed, either), I don't believe the difference is as significant as you make it

seem.  One of OS -9's most fantastic features is that it is
hardware-independant!  While I would have liked to see a single bus
architecture, any differences in functionality can be limited to the system
device drivers and the user should never have to see them.

If you're lookin' for contacts, there are lots here.  I try to support the CoCo,

but the others are right when they say good software takes years, not months.
I'm slaving away right now.  Everyday after work/school and most weekends.  Even

today when th e sun is out and hot.

                                                           Brian

-*-

30069 10-JUN 02:16 General Information
     Files Bug
     From: BRIANWHITE   To: ALL

Here is a letter I originally composed to Pete Lyall.  Unfortunately, I haven't
been able to get hold of him, so I'll post it here for everyone to read and
maybe someone else will pass it on to him...


Peter,

I came accross a very interesting bug in your "files" command the other day.  It

seems that it won't honor backup locks (even with the "-b" option) if the name
of the directory that contains the file "LCK..Backup" has been renamed to a
shorter name than it had previously.  Sounds plenty confusing, doesn't it.
Maybe an example will help.  Here's what I did to come to this conclusion.

   Directory:          GRAPHICS
   First File in Dir:  LCK..Backup

   What I Did:         rename graphics I.GRAPHICS  (for DirSort order)
                       dirsort
                       rename i.graphics GRAPHICS

                                                          _ _
   After that, the name in the root directory was "GRAPHICSCS".  I used the bar
above the letter to indicate that it's MSBit was set to show the end of the file

name.  With the directory in this state, the "LCK..Backup" (first file in the
GRAPHICS directory) was not honoured, regardless of the "-b" option.  I then
used dEd to change the trailing "CS" to null ($00) bytes.  The backup lock is
now honoured again.  The GRAPHICS directory was not the only Backup-Proof
directory that was affected by this.  Every directory in the root directory
(which had been renamed "x.DIRNAME" so I could sort in the order I wanted)
refused to honour the LCK..Backup until after I zeroed the last two chars of
each name with dEd.  Sounds like a fun bug to track down.  Best of luck. ;-)

                                                           Brian

-*-

30073 10-JUN 03:03 General Information
     RE: A new kitten in the house! (Re: Msg 29887)
     From: OS9UGPRES    To: EDDIEKUNS (NR)

Eddie - I think that since you posted your msg, that you said the kitten and the

other cat are getting along now. That's great. We still have a couple of cats
that don't like anybody <grin>.

Yah, kittens have big screams. They're also pretty fragile. We had a mother cat
once who had her kittens outside. She began bringing them in a while later, one
at a time (we couldn't figure out where she was keeping them).

The first few were dead when she brought them in. Then the last two she brought
almost together, and they were covered with fleas which had drained their blood.

Apparently they could keep going for a while, but then had simply run out of
newborn reserve energy.

Anyway, they were cold and semi-stiff, and I figured they were also dead. But
then one tried to say something, so we ran both over to the vet. He gave them a
blood transfusion (I can just picture miniature bags of blood hanging on a hook,

can't you?)... and now they're two of the fattest old cats we have! Amazing. -
kev

PS: oh yeah, be sure to pet both at the same time!

-*-

30074 10-JUN 03:03 General Information
     RE: B&B HARDDRIVE (Re: Msg 29958)
     From: OS9UGPRES    To: WILLALHUFF (NR)

Bill - is Ezgen in memory when you run the command? Or is it in a command
directory on your hard disk? - kev

-*-

30075 10-JUN 03:04 General Information
     RE: Unlink (Re: Msg 29899)
     From: OS9UGPRES    To: EDDIEKUNS (NR)

Nope. When you unlink by hand, OS-9 does what you ask <grin>... even if it's in
use. I suppose they don't go checking around because of the extra time it would
take to do so.

OSK may do it differently; I haven't checked. It does have a neat option in a
module header called the "ghost" bit (it has a more formal name, but I forget it

offhand). Oh yeah, the "sticky" bit <grin>.

That bit tells OS9 to keep it in memory when the link count goes to zero. In
practice, commands like Dir have this bit set. So the first time you type "dir",

it comes in and executes but doesn't leave. This speeds up operation to the
casual user from then on!

The other side of this is that if OS9 needs more memory for something you're
doing, then ghost modules with counts of zero are eligible to be finally
unlinked and their memory used. All in all, a pretty slick deal.

-*-

30076 10-JUN 03:04 Grits & Gravy
     RE: Interleave factor (Re: Msg 29976)
     From: OS9UGPRES    To: KENHALTER (NR)

Ken - yes, interleave is the same as skip factor.

-*-

30078 10-JUN 03:05 General Information
     RE: Questions about the MM/1 (Re: Msg 30033)
     From: OS9UGPRES    To: KEITHMARCH (NR)

Keith - I'll jump in and try some of those.

1) Is there a WEFAX/HAM program for OS-9/68K?

Not that I know of yet. I know some Hams interested in working on such tho.

2) Will the MM/1 support the RAVE package?

It could, if you can afford RAVE. In fact, I was talking with someone the other
day who's gonna port rave over, and I mentioned that if he made sure the
"display screen" call consulted the normal window CLEAR key routine, then you
could actually run my windows and RAVE at the same time! Just give RAVE its own
video ram area and screen. We chuckled at the thought of this warped idea, but
who knows?

3) What is the System ROM, How many K?

I can't recall. 64K? 128K? Something like that. OS9 itself is tiny as always, so

commands like shell, dir, etc are usually put in there also.

4) If a person does not buy the I/O Board with the Main
   CPU Board, can you use the Smartwatch RTC chip in the
   EPROM Slot used for OSK?

Interesting Q. Carl Kreider did a clock driver for the ST like that, so at least

the chances are there.

5) Is there a DMODE command for OSK?

You bet. Someone (Bob Santy?) ported it over long time ago. There may be more
than one version around, in fact. Maybe even MW has one now. Dunno.

6) Is there a terminal program that supports X/Y/ZModem?

Got me. At least XModem, I'm sure. CIS-B, for sure. Kermit, for sure.

7) Does the BASIC that comes with the MM/1 have a CLS command
   built in?

No. It wouldn't know what to send out. But of course we've done a normal command

called "cls", just as many did on the coco. Basic/68K does have inkey built in,
tho.

8) Later on, can users pull out the 68070 15 Mhz chip and put
   in a fast chip. Is the CPU socketed?

If a faster 68070 comes down the line, yes. Might need some baudrate table
changes to the main serial port (which is on the cpu chip), might not. Depends
on how Signetics makes the faster 68070... doubt they'd goof it up.

9) If I format a 3.5" 720K DS/DD Floppy Disk under OS-9/Level II
   can I read/write to it useing the MM/1 and OS-9/68K or do I have
   to reformat it to use it?

I read/write Atari ST and CoCo 3.5" disks all the time on it. Just gotta Dmode
first (or keep several descriptors around, which is what I do).

10) Will most programs use the 'F##' keys on the IBM Clone Keyboards?

Up to programmers. I suspect we'll try to push F1 as the default "Help" key tho,

at least.

11) Will the MM/1 have/be able to use, a Math Coprocessor?

Like the 68008, 68000, and 68010, the 68070 can use the math chip only as a
peripheral. Which means an expansion board or a bus board would work. Only the
68020/030/040 actually uses it as a true coprocessor.

12) Will all major chips be socketed?

Except the video chip. There's no socket made for it yet.

13) Are you/Will you be, working on a book called 'Start OS-9/68K?

He'll have to answer that one.

14) How will we change are BOOT File in the EPROM?
    a) Buy a EPROM burner.
    b) Boot off of floppy from that point onward.
    c) Send the EPROM back to you with patches on the disk
       and pay for the reburn.
    d) Pull hair out while tring to figure out what to do.

Probably all four <grin>.

-*-

30089 10-JUN 10:29 General Information
     RE: Questions about the MM/1 (Re: Msg 30033)
     From: NES          To: KEITHMARCH (NR)

Keith,  I hope I can answer some of your questons about the MM1 (2001). I dont
no the ROM size, but ROM boot's the system as far as seeing if you have a BOOT
disk or harddrive with the boot file on it, also os9 when booting will
automatical overwriten older modules with new and updated ones found on the boot

disk(or harddrive),  You need not buy a real time clock becouse there is one on
the second board, The Floppy controler will controll standered diskdrives, They
are supporting read/writening to the old type disks, As for software you get:  C

compiler, BASIC, a simple text editor, Graphics Editor, and a Terminal program,
plus OS9(K) by Microware who are supporting there 68000 version of os9. almost
forgot, You will be able to read/write to IBM disks If I missed something leave
me another message. -eric    (NES)


-*-

End of Thread.

-*-

30079 10-JUN 03:06 Programmers Den
     RE: Aciapak blues -- Chapter 2 (Re: Msg 29935)
     From: OS9UGPRES    To: EDDIEKUNS (NR)

Eddie -

Some info which might help:

1. SCF automatically calls SS.ComSt whenever an SS.Opt is done. Always.

2. SCF will call SS.Relea when a path is closed (process dies), but ONLY if:
   a) It was the last process to use that device.
   b) The process has no other paths to that device open.

3. SS.Relea only works if proc id same as the one who set the signal.

4. SCF will call SS.Close if that path desc has only one user.

5. Most serial drivers will loop on an SS.Close until all chars have been
sent.... and then shut off transmit interrupts.

It's a complicated thing, and I may have some of this wrong. But the gist is
there <grin>. - kev

-*-

30082 10-JUN 06:51 General Information
     RE: ROBOT ODYESSY I (Re: Msg 29747)
     From: WAYNELAIRD   To: OS9BERT (NR)

you lost me, what do you want to do , run it on the coco3 or put it all in your
hard drive? It'll already run on the coco3 but to run it along with levelI I
think if you bust the boot, put the modules in your startup boot for levelII set

you correct grphics window and you may take off. -w.

-*-

30083 10-JUN 06:52 Telcom
     RE: RiBBStrek (Re: Msg 29759)
     From: WAYNELAIRD   To: EASYSINGLES (NR)

thanks bob , I've got you for the upgrade, watch for it this summer. -w.

-*-

30084 10-JUN 09:49 General Information
     RE: New Guy (Re: Msg 30006)
     From: PAULRINEAR   To: EDDIEKUNS (NR)

      The line it gets stuck on is blank. Looks like the end of the file
 to me. The line before it is "}" closure to main. ???
      Paul

-*-

30085 10-JUN 09:51 General Information
     RE: New Guy (Re: Msg 30009)
     From: PAULRINEAR   To: DODGECOLT

       That's quite possible. I downloaded using X-modem. Will dump
  this file and look for junk. Thanks.
     Paul

-*-

End of Thread.

-*-

30086 10-JUN 09:53 General Information
     RE: OS9 vs. PCDOS -- gripe bucket (Re: Msg 30027)
     From: PAULRINEAR   To: KNOT1 (NR)

        OK. Any idea where that is available, 'cause I'm definitely
  interested in being impressed.
        Paul

-*-

30090 10-JUN 14:45 Grits & Gravy
     Home Publisher and printer settings
     From: POLTERGEIST  To: ALL

   I recently acquired a Star NX-1000 II at the Chicago RainbowFEST, and I am
having some trouble settig the DIP switches that control either Auto-CR or
Auto-LF.  This is my first experience with a printer of this kind.  Also, Home
Publisher has it's own Auto-LF setting.  Which setting is the correct
combination for both the printer and Home Publisher?

-*-

30093 10-JUN 15:50 General Information
     monitors
     From: JENG         To: ALL

I have a chance in buying an Amiga 1080 monitor really cheap, .42 dot pitch as
compared to .52 on the CM-8. I think I remember Kevin Darling said he used one
(1084 basically the?) I need to know what cables I need and where to buy one for

my CC3. I called Microcom and they do not think the Magnavox or Sony cables will

work. Thanks! JENG

-*-

30095 10-JUN 19:09 Telcom
     RE: Multi-port RS232 PAK (Re: Msg 29845)
     From: FILIP        To: JEVESTAL (NR)

Jim,

Thanks for the tip. I'll check that one out.

Regards,

Dave.

-*-

30096 10-JUN 19:13 New Uploads
     OSKer
     From: RAMMER       To: ALL

 *-----------------------------------------------------------------------*

      ANNOUNCING A NEW MAGAZINE FOR OS9/68000 (OSK) AND OS9 USERS

                           " the OSK'er "


 Already in the first issue (coming early July):

 o  News about the latest in OSK and OS9 Hardware
 o  A feature article on the the two new computers and their creators
 o  Information on differences between OSK and OS9
 o  An ongoing article on programming in the C language
 o  A software wishlist
 o  And many other suprises...

 Other features soon to come:

 o  Basic programming on OSK and OS9
 o  Assembly language in OSK
 o  Graphics on the MM1
 o  Software Reviews
 o  Program submissions
 o  OSK'er on Disk Subscription


 To get your own copy of the first issue FREE, or if you are convinced this
 is the magazine for you and want to subscribe, fill out the form below:
 (Due to possible response, some limits may be imposed, but unlikely)

 =======================================================================

 NAME ___________________________________      [ ] Send 1st Issue Free

 ADDR ___________________________________      [ ] Subscribe for 1 year

      ___________________________________          Include $12 in US

      ___________________________________          Or $15 Elsewhere

 MAIL TO:  the OSK'er,  P.O. Box 24285, Speedway, IN 46224

 =======================================================================

          * * * WANTED - ARTICLE WRITERS, REVIEWERS, ETC * * *

 Anyone who would like to write an article for the OSK'er, review a piece
 of software, submit a program for inclusion, or help in any way please
 contact Scott via e-mail or call (317) 241-6401.

 -----------------------------------------------------------------------

          ATTENTION VENDERS, SOFTWARE DEVELOPERS, ETC...

 Advertisements are available in the OSK'er for very competetive rates!
 Call Scott at (317) 241-6401 and ask about the special offer for the
 first issue!  Ad copy must be received before JUNE 30 for inclusion.

 *-----------------------------------------------------------------------*


-*-

30097 10-JUN 21:08 Telcom
     disable call waiting
     From: ZACKSESSIONS To: ALL

This week I am going to be getting a private telephone line as to not tie up my
parents line so much. The only custom calling feature I saw a need/use for was
call waiting, BUT ONLY WHEN NOT MODEMING!! We do live in an area which supports
"call waiting disable" and to do so requires that you dial a *70 prior to the
call. My question is, "How do I dial a * (to the left of the 0) with my modem
from my keyboard?" (I have a Hayes compatable modem, a Smart One 2400X.

Thanks!

Zack

-*-

30098 10-JUN 21:16 Device Drivers
     RE: Ramdisks & 1Meg Upgrade (Re: Msg 29988)
     From: RADARBUZZ    To: DODGECOLT

Mike, I found the VDD Driver you mentioned. I found no docs with it though, so I

did the following to try to get it to work:

1) I made a new OS9Boot file that incuded the VDD module instead of the RAMMER
module that I was using.

2) Assuming /r0 needed to be modified in some way, I first changed the name
Device Driver name at offset $26 from "RAMMER" to "VDD".  This patched /r0 is
also in the boot file.  This is a dump of it:

   Addr   0 1  2 3  4 5  6 7  8 9  A B  C D  E F  0 2 4 6 8 A C E
   ----  ---- ---- ---- ---- ---- ---- ---- ----  ----------------
   0000  87CD 002F 0021 F181 CB00 2300 26FF 07FF  .M./.!q.K.#.&...
   0010  E00F 0100 0020 0100 2801 0100 1200 1203  `.... ..(.......
   0020  0152 B052 42C6 5644 C420 2020 320F D1    .R0RBFVDD   2.Q

3) Ok, now I try to use /r0 as I have before:

   OS9:iniz /r0
   OS9:format /r0

   COLOR COMPUTER FORMATTER
   Formatting drive /r0
   y (yes) or n (no)
   Ready?  y
   Error #208

Illegal Service Request.  What's my problem?  Are there other changes that must
be made to r0? Just when I thought I had figured it out, it jumpped up and bit
me.

-Jeff


-*-

30100 10-JUN 21:23 Device Drivers
     RE: Ramdisks & 1Meg Upgrade (Re: Msg 30098)
     From: DODGECOLT    To: RADARBUZZ (NR)

The VDD type ramdisk only needs to be iniz'ed to format it. I had some
permission problems also, if I remember correctly. These can be fixed by making
sure the mode byte in the descriptor (offset 13 deccimal, $0d hex) is set to $ff

(all permissions/capabilities....)
 -Mike

-*-

End of Thread.

-*-

30099 10-JUN 21:19 Programmers Den
     C question
     From: CHYDE        To: ALL

I was just wondering if anyone else has had this problem.  When I use the
getime() system call in a program I sometimes get strange characters on the
screen or sometimes the machine locks up. I was just wondering if it was my
machine or what.  It seems to do it with both the original library and the
Kreider library.
     Chris

-*-

30101 10-JUN 23:08 Programmers Den
     RE: C question (Re: Msg 30099)
     From: ZACKSESSIONS To: CHYDE (NR)

I use the getime() function all the time in several program with no adverse
effects. Can you elaborate on the program in question?

Zack

-*-

End of Thread.

-*-


FORUM>Reply, Add, Read, "?" or Exit>