home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Cream of the Crop 1
/
Cream of the Crop 1.iso
/
PROGRAM
/
TP122092.ARJ
/
12-20-92.TPC
Wrap
Text File
|
1992-12-20
|
74KB
|
2,263 lines
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-14-92 00:43:00
From Trevor Carlsen
To Paul West
Subject Re: Fast Direct screen writess
> the only fast way to achieve this is to go into asm and
> doing the thing directly. Pascal uses the BIOS to output the
> text, so it is SLOW.
TL> Not true. Turbo Pascal, by default, writes directly to
TL> VIDRAM, bypassing BIOS writes. You can force it to write to
TL> BIOS by setting DirectVideo to FALSE.
PW> Not true.
PW> Turbo Pascal uses BIOS calls by default if you do not
PW> include the CRT unit. Including the CRT unit causes the
PW> default to change to Direct Video. You can also force it to
PW> use DOS if you wish.
This next "Not true" means we are now nested about 4 levels deep in them!
By default Turbo Pascal uses DOS to do its screen writing. If the CRT unit
is used then the default is to do direct video writes. If the CRT unit is
used and the DirectVideo typed constant is set to false then the BIOS is used.
At no time does TP use the BIOS for screen handling as a default - although
I have not checked Turbo Vision. The TP6 manual explains this quite clearly
in the second paragraph of page 199 of the Programmer's Guide.
TeeCee
--- TC-ED v2.01
* Origin: The Pilbara's Pascal Centre (+61 91 732930) (3:690/644)
* Tossed by SFToss/286 v1.02a on 92/12/16 09:14:17
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-14-92 01:01:00
From Trevor Carlsen
To Steve Mckain
Subject Record Fields to Text File
SM> The no success is that if I want to only list for instance
SM> the last name from a record file that contains first, last
SM> and age, I end up with the entire record contents in the
SM> text file, not the specific information I requested from it.
Ok, the following example will list just the LastName to the text file.
type
YourRecord = record
FirstName,
LastName : string[20];
age : byte;
end;
var
tf : text;
YRf : file of YourRecord;
YR : YourRecord;
begin
assign(tf,'TextFile.txt'); rewrite(tf);
assign(YRf,'TypedF.dat'); reset(YRf);
while not eof(YRf) do begin
read(YRf,YR);
writeln(tf,YR.LastName);
end; { while }
close(tf);
close(YRf);
end.
Untested.
TeeCee
--- TC-ED v2.01
* Origin: The Pilbara's Pascal Centre (+61 91 732930) (3:690/644)
* Tossed by SFToss/286 v1.02a on 92/12/16 09:14:17
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-14-92 23:19:00
From Trevor Carlsen
To Todd Lowe
Subject Fast Direct screen writess
> Pascal uses the BIOS to output the text, so it is
> SLOW.
TL > Not true. Turbo Pascal, by default, writes directly
TL > to VIDRAM, bypassing BIOS writes. You can force it to
TL > write to BIOS by setting DirectVideo to FALSE.
TL> Actually.. he's pretty much correct. In default, Turbo Pascal does
TL> use BIOS.
All three statements are incorrect. TP by default uses DOS and not the BIOS
for its screen handling. If the CRT unit is used then the default is direct
video writes and if DirectVideo = false the BIOS is used.
TeeCee
--- TC-ED v2.01
* Origin: The Pilbara's Pascal Centre (+61 91 732930) (3:690/644)
* Tossed by SFToss/286 v1.02a on 92/12/16 09:14:17
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-14-92 23:26:00
From Trevor Carlsen
To Vince Laurent
Subject Re: SB
VL> The code died here. COuld you repost in multiple parts?
Reposting in multiple messages is not permitted. If your reader is crippled,
get the complete message from the packet - if it is truncated in the packet,
complain to your NC as it means that someone upline from you is altering mail
- a hangable offence in FidoNet. :-)
Sorry for being so blunt about it but the plain fact is that crippled readers
and/or mailers cause the net real problems in this regard and many moderators
are now taking a stand on this.
Moderator.
--- TC-ED v2.01
* Origin: The Pilbara's Pascal Centre (+61 91 732930) (3:690/644)
* Tossed by SFToss/286 v1.02a on 92/12/16 09:14:17
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 19:50:00
From Terry Hughes
To Brandon Jones
Subject Execswap program
BJ> TH> When an swapping EXEC causes an immediate hang the problem is almost
BJ> TH> always an interrupt service routine (like keyboard or clock services) o
BJ> TH> mouse event handler left active. Be sure to turn off or reset any
BJ> TH> hardware interrupt vectors you've taken over and turn off the mouse
BJ> TH> event handler.
BJ> TH> From the way you describe the problem I'd guess the parent program left
BJ> TH> a clock interrupt service routine active.
BJ>Well I am not grabbing any interrupts directly that I know
BJ>of, but I am using a fossil in the program. Ideas now?
Using a FOSSIL driver wouldn't be a problem since that doesn't require
you to set up any interrupt handlers.
Are you using any libraries or routines you didn't write? (like Turbo
Vision or Object Professional). Are you using a mouse in your program?
Another way to gain some information about the problem would be to do a
normal non-swapping EXEC to DOS and run a program like MAPMEM to show
which interrupts your program has hooked. If you see it hooking 8 or 1C
then you know it has installed a clock ISR.
-Terry
___
X QMPro 1.0 41-2187 X TurboPower Software (voice 719-260-6641)
--- Maximus 2.01wb
* Origin: The Programmers Playhouse (1:128/60)
* Tossed by SFToss/286 v1.02a on 92/12/16 20:42:19
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 19:57:00
From Terry Hughes
To Jud Mccranie
Subject Re: TP pick problem
JM> TH> CharPickNow and CharPickSrch honored the setting of SrchStart.
JM> TH> Could you post a small test case that shows the problem you're
JM> TH> having?
JM>I extracted a sample from my program that demonstrated the problem,
JM>and while testing it I discovered something: SrchStart seems to ignore
JM>leading blanks. Knowing that, it works. I had a leading blank in the
JM>pick lines, so I thought SrchStart should be 2 (the actual column I
JM>want to search on). If it is set to 1 it works as I want it to.
The manual doesn't say it (but it ought to...) but all pick items are
trimmed before being compared. That's why the leading blanks are
ignored.
-Terry
___
X QMPro 1.0 41-2187 X TurboPower Software (voice 719-260-6641)
--- Maximus 2.01wb
* Origin: The Programmers Playhouse (1:128/60)
* Tossed by SFToss/286 v1.02a on 92/12/16 20:42:19
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 20:06:00
From Terry Hughes
To Jud Mccranie
Subject B-tree problem
JM>I have a problem in B-tree. I'm using 9 file blocks. W/o save mode,
JM>25 file handles is sufficient. with SaveMode, even 95 gives error 9901
JM>(too many open files). I have FILES set sufficiently high, and I
JM>changed ExtendFileHandles appropiately. What could be the problem?
Are you sure the call to ExtendHandles succeeded? Are you using at least
DOS 3.3? (earlier versions of DOS don't support the extend handles
call).
You can't "extend" more handles then were specified in your FILES
parameter in CONFIG.SYS. Well, DOS will let you but it will start
failing file open request when the system file table fills up. Did you
have your FILES parameter set to at least 40 or so (remembering that
other processes or TSRs might be using up slots in the system file
table).
The bottom line is that an appropriately sized FILES parameter and a
call to ExtendHandles is all that's required. If you're confident you
are doing those things right then it's time to make a little test
program (outside of Filer) to see what happens there.
-Terry
___
X QMPro 1.0 41-2187 X TurboPower Software (voice 719-260-6641)
--- Maximus 2.01wb
* Origin: The Programmers Playhouse (1:128/60)
* Tossed by SFToss/286 v1.02a on 92/12/16 20:42:20
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 00:00:00
From Terry Hughes
To Jakob Paikin
Subject TPErrHan and protected m
JP>Hello Terry!
JP>I was looking at the makefile for TPRO 5.20, and saw that
JP>TPErrHan wasn't compiled for protected mode.
JP>As I can't really figure out if TPErrHan is capable of
JP>working in protected mode or not, I thought it better to
JP>ask you.... What do you say?
TPERRHAN was updated to work in protected mode. I'd have to say it was
just an oversight that it wasn't included in the protected mode
portion of the make file. Thanks for pointing that out.
-Terry
___
X QMPro 1.0 41-2187 X TurboPower Software (voice 719-260-6641)
--- Maximus 2.01wb
* Origin: The Programmers Playhouse (1:128/60)
* Tossed by SFToss/286 v1.02a on 92/12/16 20:42:20
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 00:00:00
From Terry Hughes
To Jud Mccranie
Subject Tricky bug in B-tree
JM>test cases. So if anyone has a problem where it just won't find what
JM>they're sure is there, ask if they are doing something like this.
Thanks for the note. We have a standard litany of "things you might
be doing wrong during key generation" when people call with such
technical support questions. I'll add that one to the list.
-Terry
___
X QMPro 1.0 41-2187 X TurboPower Software (voice 719-260-6641)
--- Maximus 2.01wb
* Origin: The Programmers Playhouse (1:128/60)
* Tossed by SFToss/286 v1.02a on 92/12/16 20:42:20
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 00:00:00
From Terry Hughes
To Douglas Delonay
Subject lock
DD> I changed it to accept a untyped variable as the file parameter. when
DD>passing it an untyped file I would ned to pass it the place
DD>to start locking in bytes? i.e pass it
DD>filepos(f)*sizeof(record_size) for the filepos to start
DD>locking at? Thanks for any help. Take care!
Richard Sadowsky doesn't monitor this echo. I don't have his LOCK4
unit here at home but I can tell you that your assumption is correct:
if you pass in a typed file the byte offset to lock would be as you
specify above.
-Terry
___
X QMPro 1.0 41-2187 X TurboPower Software (voice 719-260-6641)
--- Maximus 2.01wb
* Origin: The Programmers Playhouse (1:128/60)
* Tossed by SFToss/286 v1.02a on 92/12/16 20:42:20
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-14-92 07:42:00
From Dj Murdoch
To Nathan Wambach
Subject Re: pascal uncompileer
NW> Hmm, You are right, of course, but I was just saying... that it
NW> could be done.. Anyway, I am wanting to learn ASM... got any names for
NW> GOOD books to read??? Thanks!
The Intel manuals and the manual that came with A86 are all that I ever read.
--- Msg V3.2
* Origin: Murdoch's_Point - - (1:249/99.5)
* Tossed by SFToss/286 v1.02a on 92/12/16 20:42:23
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 21:25:00
From Trevor Carlsen
To All
Subject 15/12/92 - contest update
Well, my BP7 arrived yesterday so I will recompile my program and get it out ASAP.
To date there has only been one entry (thanks Bob S.) Running time was approxim
tely 84 minutes. (My TPTimer timing unit from Turbo Power is not usable after
60 minutes, so the above assumes a 60 minutes rollover.)
Bob S, I know you indicated that this is a preliminary effort with a great
deal of optimising still to go, but it still introduced some interesting concept
! Good one! Just one point you should note... in order to run on an XT it
will require the G+ switch changed to G-.
At the moment the time to beat (TeeCee's) is 909.432 seconds (TP6) however
that should come down a little with TP7. Comparisons for judging will all
be compiled under the same compiler but as authors you determine the compiler's
switches and settings. I will not change that.
TeeCee
--- TC-ED v2.01
* Origin: The Pilbara's Pascal Centre (+61 91 732930) (3:690/644)
* Tossed by SFToss/286 v1.02a on 92/12/16 20:42:43
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 07:00:00
From Trevor Carlsen
To Steve Connet
Subject Pointers
SC> It is difficult for me to understand linked lists. I seem to be able
SC> to
SC> go "forward" easily, but how does one go "backward?" For example:
SC>
SC> Type CurrentPtr = ^EachRecord;
SC> EachRecord = Record
SC> Dummy : Integer;
SC> Link : CurrentPtr
SC> End;
SC>
SC> Var CurrentRec,
RootRec : CurrentPtr;
SC>
SC> Procedure GetNumber;
SC> Begin
SC> With CurrentRec do
SC> Readln(Dummy)
SC> End;
SC>
SC> Procedure ShowNumber;
SC> Begin
SC> With CurrentRec do
SC> Writeln(Dummy)
SC> End;
SC>
SC> Begin
SC> New(CurrentRec); { initialize new record }
RootRec := CurrentRec;
SC> Write('Enter first number: ');
SC> GetNumber;
Now here is where you seem to derail...
SC> CurrentRec := CurrentRec^.Link; { advance to next record, right? }
No... the first line has already given CurrentRec its value. To create a new
record and give Link its value...
new(CurrentRec^.Link);
CurrentRec := CurrentRec^.Link;
Then scrap this next line.
SC> New(CurrentRec); { initialize new record }
With any linked list it is important to keep a record of just whereabouts
the root record is, otherwise it is lost. In the above example you do not
do this so you cannot go back to the root entry. Also, as the above is a "singly
linked list" (SLL) you cannot "go backwards" from the current point. Providing
you have kept a record of where the root is, a SLL must always be read/accessed
sequentially from root to current. To move in either direction within a list
requires a "doubly linked list" (DLL) which involves two pointers within the
record - one to the previous record and one to the next record.
I have made some minor changes to the above code (the unquoted lines) that
will help you get started.
TeeCee
--- TC-ED v2.01
* Origin: The Pilbara's Pascal Centre (+61 91 732930) (3:690/644)
* Tossed by SFToss/286 v1.02a on 92/12/16 20:42:43
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-14-92 18:16:00
From Dj Murdoch
To Wayne Boyd
Subject Re: Opening for reading
WB> Anyone know the methodology for opening a file for
WB> read-only which is already in use by another program? As
WB> an example, opening the nodelist to scan or read, though
WB> in use by Frontdoor, or whatever.
The ability to do this depends on whether SHARE is loaded, and whether the
other program wants you to be able to open the file.
The idea is to set the Filemode variable properly before you try to open the
file. A good value to try is 0; that's "compatibility mode, read-only access";
another one is $40, which is "deny none mode, read-only access". Look in
a good DOS reference for a discussion of all the other possible values. I
don't know how to determine which value to use in your particular case other
than by trial and error.
--- Msg V3.2
* Origin: Murdoch's_Point - - (1:249/99.5)
* Tossed by SFToss/286 v1.02a on 92/12/16 20:42:49
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 08:19:00
From Dj Murdoch
To Max Maischein
Subject Re: Cold Boot in program: ho
> procedure Reboot; far; assembler;
> asm
> xor ax, ax
> mov ds, ax
> mov [0472h], ax
> dec ax
> push ax
> push ds
> end;
MM> Well, this reboot routine seems to jump to the wrong
MM> location ( it returns to 0000:0000, where a 'normal'
MM> reBoot routine jumps to FFFF:0000. I don't know if this
MM> was an oversight on your part, but this will only work in
MM> some cases, in others it might not !
Take another look, or watch the registers as you trace through it. It works
as advertised.
--- Msg V3.2
* Origin: Murdoch's_Point - - (1:249/99.5)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:52
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 08:23:00
From Dj Murdoch
To Ping Hansen
Subject Re: One more bugfix-message :-(
PH> I have a suggestion for you. Since it is evident that the
PH> new TP is bug ridden (as usual),
Hmmm, why would you say that? I've only seen two serious bugs so far (SHL/SHR
for longints, and the file dialog). There are a number of glitches in the
IDE, but those hardly matter at all.
PH> then why don't you put
PH> together a masterlist and post it every time a new error
PH> surfaces or once every week during the initial phase - say
PH> until March 1st ? I'd rather have the increased bandwidth
PH> than have my programs go beserk :-(
I'll probably put together a BP7 bug list, but it takes time, and I'm not
likely to have any for a few weeks. I won't post the complete list here,
but I'll post updates.
--- Msg V3.2
* Origin: Murdoch's_Point - - (1:249/99.5)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:52
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 08:28:00
From Dj Murdoch
To Paul West
Subject Re: Debug data format / Map file convers
JR> I am surprised Borland did not foist this one on you when you ordered JR>
BP7. They sold me two add-ons at about US$50 each when I phoned in my JR>
order. First was PROTOGEN, an outside designed Windows source code JR> generator
for menus.
PW> I'm going to take a close look at my credit card slips
PW> when they come in. I was promised protogen FREE with my
PW> BP upgrade if I paid the $5.00 shipping charge. If they
PW> charge me 50 bucks for it, I'm going to scream bloody murder.
From what I've heard, it was free to people who had both TP and TPW, $50 to
people with just one of them.
--- Msg V3.2
* Origin: Murdoch's_Point - - (1:249/99.5)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:52
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 08:33:00
From Dj Murdoch
To Brian Pape
Subject Re: Bp 7.0 price
BP> Does anyone out there think that there is any way to
BP> complain to Borland about the price I got charged on
BP> BP7.0? I called them to order the day I got their upgrade
BP> postcard, and the guy told me it would be $159.95 (+/- 5
BP> cents) including shipping and handling... When the
BP> Mastercard bill arrived, they had charged me about $185.
BP> This is especially annoying because it is cheaper than
BP> this MAIL-ORDER for the upgrade (Telemart & PC-zone),
BP> plus, since I'm going to school, I don't make enough money
BP> to pay for their misquotes...
What did the invoice included with the package say? Generally it itemizes
the charges. What did customer service at Borland say? I've found them to
be very reasonable about things like this.
--- Msg V3.2
* Origin: Murdoch's_Point - - (1:249/99.5)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:52
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 08:35:00
From Dj Murdoch
To Brandon Jones
Subject Re: BBS
EG> #2: Does turbo automatically close OPEN files after an Exit or Halt?
BJ> Far as I know, yes.
No, only Input and Output are closed by TP. For other files it's up to you
to call close. DOS will close everything, but if you didn't close it through
TP, you'll lose anything that's sitting in a TP buffer.
--- Msg V3.2
* Origin: Murdoch's_Point - - (1:249/99.5)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:52
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 19:29:00
From Dj Murdoch
To Jud Mccranie
Subject Re: BP7 ide bug
DM> That's the the way it's designed to work - the file open window
DM> does the same thing. Hitting Enter opens the file that's
DM> listed in the input line, not the highlighted one below.
JM> But it DOESN'T work right. I still believe that it is either a bug
JM> or a major design flaw.
Is that part you quoted all of my post that you read? I'll repeat: TAB is
used for navigation. The only way to change the focus to the buttons is to
hit TAB repeatedly. Would you want moving the cursor to a button to have
the side effect of changing the primary file?
--- Msg V3.2
* Origin: Murdoch's_Point - - (1:249/99.5)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:52
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 19:40:00
From Dj Murdoch
To Nathan Wambach
Subject Re: exe's
NW> OK, a program needs to know what it's own EXE file is called,
NW> how do you do it? What I wanted to use this for is linking files into
NW> the EXE... You have to know the EXE file name and who knows, maybe
NW> someone renamed it???
As long as you're in DOS 3+, the filename is available in paramstr(0). If
the program was run from COMMAND.COM, it'll have the full path to the file;
if it's run from another program by an Exec call, it'll have whatever you
give as the filename parameter.
--- Msg V3.2
* Origin: Murdoch's_Point - - (1:249/99.5)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:52
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 05:45:00
From Mark Ouellet
To Paul Wilden
Subject Re: ARJ EXTRACTOR
On 03-Dec-92, you, Paul Wilden, of 1:215/208.0 wrote...
PW> Use FindNext to the next file matching the same attributes as the call
PW> to FindFirst.
Thanks Paul,
but I allready knew that!!! What did I say to generate that reply
though ???
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:52
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 05:47:00
From Mark Ouellet
To Christopher Turcksin
Subject Re: Turbo Vision menus
On 05-Dec-92, you, Christopher Turcksin, of 2:292/600.21617 wrote...
CT> Would I be waiting long for the TP7 version book? I think I saw it in
CT> the book-store, but after I was searching something specific on
CT> TurboVision ...
Well Christopher,
it depends, if Mr Rubenking was part of the TP 7.0 Beta testers then
he might have an update comming very soon. If on the other hand he only
got his copy a few weeks ago like every one of us then it might be a
little bit longer. There also is no warranty that a TP 7.0 version is
forthcomming. That's the problem with these authors, you never know if
there is going to be a re-edition until it actually comes out.
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:52
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 05:54:00
From Mark Ouellet
To Scott C Virtes
Subject Re: Perfect Numbers, Primes, and weird n
On 02-Dec-92, you, Scott C Virtes, of 1:202/1207.0 wrote...
SC> Some of these algorithms may interest the programmers in the
SC> audience ...
Well Scott,
One I allways found interresting was the formula found by an Indian
graduate from Oxford (I think).
His formula calculated the number of UNIQUE additions to form any
number.
Like this:
2 = 0 + 2
= 1 + 1
3 = 0 + 3
= 1 + 2
= 1 + 1 + 1
4 = 0 + 4
= 1 + 3
= 1 + 1 + 2
= 1 + 1 + 1 + 1
= 2 + 2
5 = 0 + 5
= 1 + 4
= 1 + 1 + 3
= 1 + 1 + 1 + 2
= 1 + 1 + 1 + 1 + 1
= 1 + 2 + 2
= 2 + 3
I might have forgotten some possible combinations but you get the idea!!
What was most extraordinary of his formula is that it actually gave
a REAL number with a fractionnal part BUT, Truncing the result gave the
exact number of combinations possible for any number.
I don't recall what the formula was but maybe you do!!??
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:53
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 07:29:00
From Mark Ouellet
To Greg Vigneault
Subject RE: LONGINT SHIFT BUG
On 04-Dec-92, you, Greg Vigneault, of 1:250/304.0 wrote...
GV> I didn't expect that. Did SHR31 work on your '486? If other '386
GV> and '486 users try SHR31, it should let them know if they will
GV> experience problems with TP7. If it's inappropriate to post such
GV> information here, I'd be interested in having it netmailed (which
GV> systems failed to shift the 1 through all 32 bits).
Greg,
It worked fine on my 486-DX33 [Except for the fact I went through
more than one run and I ended up running a 986 DX|SX then a :86 then ;86
etc... ;-)]
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:53
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 07:43:00
From Mark Ouellet
To Dj Murdoch
Subject Re: BP 7 bug
On 04-Dec-92, you, Dj Murdoch, of 1:249/99.5 wrote...
JP>> The only DOS-related (I haven't installed the
JP>> Windows-part) files that are different from the distribution-files
JP>> are:
JP>> \BP\BIN:
JP>> TURBO.TPL
JP>> TPP.TPL
JP>> TPW.TPL
JP>> \BP\UNITS:
JP>> DRIVERS.TPP
DM> I don't understand this. What are you comparing here?
Dj,
Maybe his comparing the BETA and official releases!!
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:53
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 07:45:00
From Mark Ouellet
To Dj Murdoch
Subject Re: BP 7 slash and burn
On 04-Dec-92, you, Dj Murdoch, of 1:249/99.5 wrote...
DM> It's probably not a good idea to skip the example files. There's some
DM> nice code there; for example, the TVDebug unit and the Gadgets unit are
DM> both in the examples directories.
Dj,
I personnaly liked the TV & OWL Chess demos but specially the TV
File Manager!!!!
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:53
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 16:48:00
From Mark Ouellet
To Brian Swanson
Subject Re: Turning off screen
On 06-Dec-92, you, Brian Swanson, of 1:123/419.0 wrote...
>> SwapVectors;
>> Exec (GetEnv('COMSPEC'),'/C CALLED >Nul');
>> SwapVectors; ^^ ^^^^
BS> Nope...I tried this already...The program still writes to the screen....
Brian,
In this particular case, would CALLED refer to a TP Compiled program
too??? 'Cause if that is the case then it means, you would have to
recompile it with a small fix to allow redirection.
If it's not TP compiled then it could also mean the program writes
it's output to StandardError.
You see there are basically two devices you can use to write to the
screen, two defined by DOS that is. The "C" Language calls them
"Standard Output" and "Standard Error". The latter being normally
reserved for error messages. It purpose is to give the programer a
channel where he can be sure the user sees the error message when it is
printed. For this same reason, it is not redirectable by DOS. It can be
done but with much more work under DOS than under UNIX or most other
OSes for that matter.
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:53
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 22:27:00
From Mark Ouellet
To Evan Croskery
Subject Re: Programing
On 08-Dec-92, you, Evan Croskery, of 1:163/416.0 wrote...
EC> What's the echo tag for the "beginers Pascal" echo?
PASCAL_LESSONS
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:53
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-15-92 23:21:00
From Mark Ouellet
To Brian Pape
Subject Re: Far keyword
On 08-Dec-92, you, Brian Pape, of 1:2250/1.0 wrote...
BP> (I tried all 6 combinations of the three keywords to make sure It wasn't
BP> just looking for something it couldn't find :)
BP> anyone know why it won't accept this?
Brian,
Seems to me you might be looking too FAR for something you allready
got!!!
If I'm not mistaken, an Interrupt procedure call is preceded by the
CS:IP and Flags being pushed onto the stack before entering the
interrupt routine.
Seems to me CS would be Segment and IP offset, now doesn't that look
like a far call to you.
Maybe your reason for trying to FORCE a far attibute on the
procedure is to be able to pop the flags and do a FAR return BEFORE the
normal IRET. In which case, I sure hope TP would take into account that
it is inside an Interrupt procedure and make that a FAR return.
I assume the interrupt is automatically a FAR procedure with the
added effect of getting the registers passed to it as required by the
parameter list.
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:53
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 00:32:00
From Mark Ouellet
To Benjamin Schollnick
Subject Re: Printing....TABSTOPS
On 09-Dec-92, you, Benjamin Schollnick, of 1:2613/477.0 wrote...
BS> If I print from the *IDE* this is how it looks:
BS> Suggestions?
Benjamin,
You are, as most of us are, using TAB settings other than the
standard 8 caracters. You are probably also instructing the IDE to USE
Tabs in the code so that it alligns correctly when you indent/unindent.
Problem is those tabs are probably sent to the printer VERBOSE, tabs
are by default 8 caracters wide on printers.
So something like:
Each "+" represents a tab position if TABs are defined at 2 in IDE.
-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
Procedure A;
^--- Tab here (Only one)
Could very well expand to this:
On printer each tab is at a position which is a multiple of 8
-------+-------+-------+-------+
Procedure A; on a printer
^_____ Again only One tab but the next TAB happens to be a
lot farther then expected
hope this helps. I'll see if there is a way to prevent this:
Note that *I* Usually use WP or any available wordprocessor to print my
code. This way I can also select a print-pitch such as a 20 cpi font
which still gives me over 132 columns on plain 8 1/2" paper.
One last note, if you have BP 7.0 and have windows installed, you might
try using the Windows version of the IDE to print your code. Since it
normally leaves all the printer handling to Window's printer driver you
might get better results. Maybe even retain the "Highlighting" by using
Italics and such different print styles when printing. I have't tried it
yet myself but it might work.
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:53
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 01:10:00
From Mark Ouellet
To Benjamin Schollnick
Subject Re: Printing....TABSTOPS
Well Benjamin,
I just went and tried what I mentionned in my previous message.
FOR YOU AND ALL OTHERS
THE WINDOWS VERSION OF BP 7.0 WILL IN FACT USE COLOR HIGHLIGHTING
WHEN PRINTING IF YOU INSTRUCT IT TO DO SO (PROVIDING YOUR PRINTER
CAN PRINT IN COLOR OF COURSE) In fact it is even smart enough to use
a font that will prevent word-wrapping of your code and comments. It
probably simply used a font that allowd at least 132 columns wide
printing. I didn't print the whole thing because on my printer that
would have been a waste of my color cartridge (DJ 500C blends the 3
primary colors to produce black and since black is the predominant
color in printouts ;-)...)
I'm now thinking of selling my brand new 500C and upgrade to a 550C
which uses the new 4 Ink cartridges (Black/Yellow/Cyan/Blue). It
should waste a lot less ;-)
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:53
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 01:17:00
From Mark Ouellet
To All
Subject Highlighted printing in BP 7.0
Hi All,
In case you haven't seen the personnal reply to Benjamin,
The windows version of the BP 7.0 IDE ( BPW ) will use COLOR
highlighting to print your source if you instruct it to do so and
your printer is capable of color.
Of course you need to have your printer's driver available for
Windows 3.1 for it to use the full capability of your printer.
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:53
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 01:22:00
From Mark Ouellet
To Walter Trocquet
Subject Re: Quattro Pro Passwords
On 10-Dec-92, you, Walter Trocquet, of 1:396/5.0 wrote...
WT> Does anyone have any Pascal source code or suggestions on how to decode
WT> or simply remove passwords from Quattro Pro or Lotus spreadsheets. We
WT> have a number of spreadsheets that were password protected by a former
WT> employee and need to be able to access them. Any help or suggestions is
WT> appreciated.........Thanx, WAT, Jr.
Walter,
If you do find the answer I'd like to know. Lotus will probably be
easier, provided it is an old version. Newer versions will probably use
a more efficient encryption scheme.
Quattro Pro will be the hardest I would think. Even in the days of
SuperKey (Remember that keyboard macro program), Borland was allready
offering DES encription capability. So my guess is Quattro Pro will be
even better.
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:53
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 01:27:00
From Mark Ouellet
To Richard Nelson
Subject Re: Pointers
Richard,
I hit a wall while trying to fiddle with redirection and such stuff
in BP 7.0.
I basically declared a pointer to an Array [0..0] of <some type>
When I tried assigning it a value:
type
FHT = Array[0..0] of byte;
FHTPtr = ^FHT;
var
MyFHT : FHTPtr;
begin
GetMem(MyFht, 255); {This should reserve enough space for 255
elements}
MyFht^[2] := 3; { BP returns with CONSTANT OUT OF RANGE }
I disabled ALL compiler directives and it still gives me that same
error!!!
ANY IDEAS????
BTW, I finally tried using BPW to print my source and it does indeed
use color with my DJ 500C if I ask it to!!! It also automatically
selects a font that gives me 132 Column printing (At least 132).
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:53
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 01:40:00
From Mark Ouellet
To Christian Brem
Subject Re: BP 7.0 and Turbo Vision
On 09-Dec-92, you, Christian Brem, of 2:310/3.16 wrote...
CB> I have heard, that in TP 7.0 there will be no Turbo Vision but something
CB> similar and compatible to Windows (?).
CB> Is this true -have I to rewrite all Programms developed in TP 6.0 using
CB> Turbo Vision ?
No Christian,
If you get TP 7.0 there is still TurboVision in it only it's now up
to Version 2.0. Some minor additions to it, added functionality such as
input routines if I'm not mistaken.
However, if you get BP 7.0, you get TurboVision 2.0 again but you
also get OWL (Object Window Library) A windows counterpart to
TurboVision of some sort.
Also note BP 7.0 comes with the Tasm assembler Ver 3.2, TD Turbo
debugger, Turbo profiler (Borland's, not turbo power's profiler). It can
compile for Standard DOS, Protected mode DOS or Windows.
TP 7.0 only compiles for standard DOS. And you don't get the
assembler, debugger etc... ONLY TURBO Pascal. In fact BP 7.0 contains
every thing TP 7.0 does plus a lot more.
Oh and BP 7.0 comes with Source to ALLMOST everything. The only source
not given are those of the co-pro emulator and of the overlay manager I
think.
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:53
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 02:05:00
From Mark Ouellet
To Dj Murdoch
Subject Re: Debug data format / Map file convers
On 09-Dec-92, you, Dj Murdoch, of 1:249/99.5 wrote...
DM> It's described in the "Open Architecture" book that came with (or was
DM> optional with) BC++ 3.1. I've heard that a BP specific version is
DM> planned, but I haven't heard a publication date.
Dj,
I think it is allready available.. I got the offer with my BP 7.0.
$49 for it I think.
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:53
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 02:25:00
From Mark Ouellet
To Dave Speringo
Subject Re: DROPPing to DOS over a phone line
On 10-Dec-92, you, Dave Speringo, of 1:141/333.0 wrote...
DS> Well, I'm programming using the routines supplied by Jpdoor, so I don't
DS> think I would need Doorway, but thanx.
DS> ALL: I've used the exec to a batch but it all locks up in a batch and I
DS> don't understand! Does it have soemthing to do with my memory statement
DS> of...{$M $8014,0,40000}? Help?
Dave,
You can EXEC .EXE and .COM files only. To exec a BATCH you actually
EXEC Command.com and give the batch as a parameter.
Like this:
EXEC(GetEnv('COMSPEC'), '/c <name of batch>');
^^^^^^^^^^^^^^^^^ ^^
^ ^ this tells command.com to execute what
^ ever command follows and to return as
^ soon as it is finished (Instead of having
^ to type exit to get back)
^
This will make sure the correct command process is executed. If
you were running 4DOS then it would call 4dos.com by retreiving
the comspec variable from the environement.
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:53
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 02:41:00
From Mark Ouellet
To Norman Yen
Subject Re: vga programming/ vector things
On 10-Dec-92, you, Norman Yen, of 1:153/900.0 wrote...
NY> If anyone should do any
NY> freqing I would like to ask that you leave me some netmail. I've already
NY> had a few people freq files from my board without any messages or
NY> anything. I like to know who is doing what. :)
Norman,
I'm probably one of those people who frequed your board. Can't
remember if I had the thank-you note activated in Yabom though.
It's kind of hard to know, some sysops find it annoying and that it
only clutters their Netmail area while others, such as yourself, expect
to receive some.
I think the Nodelist should include a flag so that FRequester
programs such as Yabom, Amax etc.. can decide wether a message should be
posted or not ;-)
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:53
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 02:51:00
From Mark Ouellet
To Paul Wilden
Subject Re: PASSING PARAMETERS
On 10-Dec-92, you, Paul Wilden, of 1:215/208.0 wrote...
PW> I have question about passing variable parameters to a unit.
PW>
PW> procedure [name](Param1: type; Param2: type; var Param3: type; Param4:
PW> type);
PW>
PW> It was my understanding that with this example Param1 and Param2 would
PW> be value Parameters and Param3 and Param4 would be variable Parameters.
PW>
PW> However when I tried this in a unit Param4 was treated as a value
PW> Parameter, I had to put a second var in front of Param4 to get it to
PW> work. Can anybody explain why this is?
Paul,
You have to specify "VAR" for each variable parameter.
In your example however, if Param3 and Param4 were of the same type you
could:
Var Param3, Param4 : type;
This way "Var" and "Type" would apply to both parameters.
If they are different types then you must
VAR Param3 : type; VAR Param4 : type;
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:53
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 03:00:00
From Mark Ouellet
To David Solly
Subject Re: Printer woes!
On 11-Dec-92, you, David Solly, of 1:163/215.0 wrote...
DS> Is there a way to convert my lookup program into a TSR program, (Someone
DS> suggested going the IRS route. I am not sure what that means. Maybe
DS> someone could explain?), that will work in conjunction with the word
DS> processor and the font generator? If so, could someone, please, post a
DS> skeleton program I can use for a model?
Well David,
I doubt that the IRS (Internal Revenue Service) will be of any help ;-)
However, an ISR (Interrupt Service Routine) would help a lot. The
Basic difference between an ISR and a TSR is:
A TSR HOOKS up an ISR onto some existing interrupt vector and exits,
leaving code in memory (The ISR) to service the interrupts.
You could do it without resorting to a TSR.
write a program that:
1: Hooks ISR to an interrupt (Save the old vector)
2: EXEC your word processor
3: Un-Hooks the ISR (Restoring the old vector)
4: Return to DOS.
This way you are not losing precious memory all the time for that
conversion routine. You only load it when working with your
WordProcessor. You will have to decide what interrupt to hook in step 1
though. If the wordprocessor sends data to the printer through an
interrupt (Like a BBS sends data to a comm port by calling an interrupt)
then you could hook your ISR to that interrupt. This way it could scan
caracters sent to it and decide wether to act on them or to send them
through without modification.
Hope this clarifies it for you and that it helped in some way.
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:54
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 03:47:00
From Mark Ouellet
To Mike Copeland
Subject Re: borland pascal 7.0
On 11-Dec-92, you, Mike Copeland, of 1:114/18.10 wrote...
MC> Dj, I just received my bundle of BP7.0 - it's HUGE!!
MC> Question I'm not very sure on, and obviously haven't had enough time
MC> to research: is TP7.0 a subset of BP7.0, and do I now have TP7.0? (I
MC> really don't expect to use most of the new stuff: Windoze development,
MC> DLLs (whatever they are...), OOP, etc. - I just wanted a new, faster,
MC> smaller .EXE-producing TP...) Thanks... 8<}}
Mike,
Yes BP 7.0 is Basically TP 7.0 + the added stuff ie: Tasm, TD, TF etc...
If you don't intend to compile for Windows or Protected mode then
just use TURBO.EXE, that's the TP 7.0, Standard DOS only version.
BP is the DOS version of the IDE that compiles for DOS, PM and Win
BPW is windows version of above.
BPC is DOS command line version of BP
TPC is command line version of TURBO.exe
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:54
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 04:11:00
From Mark Ouellet
To Bruce Ruona
Subject Re: Protected Mode IDE problems
On 11-Dec-92, you, Bruce Ruona, of 1:2280/1.0 wrote...
BR> I seem to be experiencing some amount of problems related to the new
BR> protected mode IDE (BP.EXE)....
BR>
BR> Holding down either the UP/DOWN GREY arrow keys while passing through a
BR> screen or two of code, will result in an errant '2' or '8' being left
BR> scatterred throughout the code I've passed through, this will leave
BR> about two or three of these per 25 line screen-repeated presses/releases
BR> of the key does NOT appear to result in this problem.
Bruce,
This seems more like a problem with your keyboard than with BP.
Extended keyboards, specifically the "Additionnal" Pgup Pgdn etc..
section is a bit tweeked. To allow you access to those keys, the
keyboard is essentially sending codes to tell the computer:
Numlock is off
User pressed pgup key
Numlock is back on
Of course each of these codes represents MANY values by them selfs.
Each being sent as Make/Break codes. Make codes are when you press down
on the key, break when you release.
The keyboard is essentially simulating the pressing of multiple keys
so that the end result appears as PGup or PgDn etc...
I've seen this problem mainly on 386 computers. The problem seems to
be in the timing. The computer loses some of those codes so that some
times it doesn't see the code that tells it the keypad is in numlock
state and thus interprets the key as "9" instead of the intended PgUp.
Most often the result was that Numeric keys, those on the top row of
the keyboard appear to remain in a shifted state which a few presses of
the Shift key can correct by "re-synchronizing" the keyboard and the
keyboard micro-controller on your motherboard.
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:54
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 04:27:00
From Mark Ouellet
To Jud Mccranie
Subject Re: borland pascal 7.0
On 10-Dec-92, you, Jud Mccranie, of 1:3645/20.0 wrote...
JM> The old versions refused to run in 386 mode if QEMM was running. I got
JM> BP 7 yesterday, but I haven't tried the new TD or TProf yet.
Jud,
They still can't run concurently with Qemm. It is predictable since
they rely on running in protected mode and require having the highest
priority in the system.
I must admit I have tried loading TD386.exe but I haven't tried
installing TD386.sys BEFORE Qemm in my config. Maybe that could work.
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:54
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 04:36:00
From Mark Ouellet
To Jud Mccranie
Subject Re: BP7 ide bug
On 11-Dec-92, you, Jud Mccranie, of 1:3645/20.0 wrote...
JM> I believe there is a bug in the BP7 IDE (and presumably in TP7 as
JM> well). It involves selecting a new primary file when the new file
JM> you want is the *first* one in the window. If you have a primary
JM> file selected and go to Compiler/Primary, hit tab to go to the file
JM> window, it highlights the first entry and gives info about it.
JM> However, if you press ENTER to select it at this point, the Primary
JM> file DOES NOT CHANGE. If you use the arrow to move off of the first
JM> one and move back and then press ENTER it does select it. So it
JM> seems like you have to move off the first one and come back to it
JM> in order to select it.
Jud,
I tried it and you are correct. However hitting TAB to move to the
file-list window, then SPACE does update the primary file selection.
I guess the programer in charge of that bit of code never change the
primary file to the first one listed ;-)
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:54
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 04:43:00
From Mark Ouellet
To Dj Murdoch
Subject Re: Array of Pointers
On 10-Dec-92, you, Dj Murdoch, of 1:249/99.5 wrote...
DM> Actually, TP6+ uses mod 8 chunks - but the effect is the same. Seven
DM> bytes get wasted in this allocation. However, there's a big advantage
DM> to the TP 3/6+ scheme over the 4/5.x scheme. A deallocation is
DM> guaranteed to increase free space, and it's guaranteed to succeed. If
DM> you call Dispose in 4/5.x and create a new block of free heap space, the
DM> heap manager has to allocate a new 8 byte record to describe it. If it
DM> can't do that, you get a "heap overflow" error on a *de*allocation. In
DM> the current scheme, the record is stored in the space that was just
DM> freed up, which is guaranteed to be at least 8 bytes.
Dj,
Adjacent FREE memory blocks are also merged back into a larger one.
Avoiding the possible problem of having 400K free but no blocks large
enough to allocate a 1K area.
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:54
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 05:01:00
From Mark Ouellet
To jim barchuk
Subject Re: tvision echo
On 11-Dec-92, you, jim barchuk, of 1:141/1156.0 wrote...
JB> hi all.
JB> the turbovision echo is getting off the ground and is in search of new
JB> users.
JB>
JB> these nodes are actively carrying the tvision echo.
JB> dial in as a user or ask for a feed via netmail.
JB>
JB> bill himmelstoss 112/57 the dragons domain 904-268-9203 fl 24
JB> dennis powers 141/1135 uncanny-x board 203-879-7122 ct 96
JB> jim barchuk 141/1156 pontoon 203-757-7591 ct 24
JB> bill arlofski 142/111 reverse polarity 203-620-0182 ct
JB> peg lagier 208/2 turbocity 209-599-7435 ca
JB> richard nelson 216/117 clever sheep snubbs 408-458-5268 ca 96
JB> billie cohen 3604/15 land of the lost 601-467-0801 ms 96
JB>
JB> and if i'm reading the seen-bys correctly the echo should also be
JB> available to
JB> users at these nodes.
JB>
JB> mike gould 112/36 dba software solutions 904-272-5915 fl
JB> john jameison 216/506 mountain retreat 408-335-4595 ca
JB> laurent breton 240/20 lab bbs 418-648-6621 on
^^
|-------------------|
That should be Qc for
Quebec City
Province of Quebec
Canada.
Jim, in NET 240 same city/province & Country you can also add:
Douglas Kitson 240/1 [SQUARE-HEADs] 418-523-3117 Qc
Andre Morin 240/507 Bab-O-Manie Node 1 418-648-9590 Qc
Andre Morin 240/508 Bab-O-Manie Node 2 418-648-0691 Qc
JB> i have packets on hold for a couple of other folks. i'll crash it out
JB> to them and see if they're still up for it.
JB>
JB> also
JB>
JB> berend de boer 2:281/527.23 nd is hooking up now to be our feed to
JB> europe. if you're on that side of the pond and you're interested, ask
JB> berend.
JB>
JB> ALL replies and discussion about this should be via NETMAIL, please.
jim,
I replied here as the new numbers and boards could be of interrest
to other Pascal Echo participants from Net 240 who might be unaware of
the TurboVision echo's availability.
Participants or sysops from nets close to mine could also be
interrested to hook up to one of these.
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:54
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 16:08:00
From Mark Ouellet
To All
Subject BP 7.0 Utilities (Borland Canada)
Hi All,
In case this can help someone.
I just ordered
ProtoGen for Pascal $49.95 (Maybe US price, I'm not sure)
&
Open Architecture Handbook for BP 7.0 $59.95
Total Cost was $161 including $10 shipping/product and I also
suspect that Provincial and Goods & Services Tax are included in
that price.
The Windows API Reference for Pascal is not yet available from
Borland Canada as of Wednesday, December 16th 1992. It should be
available in a couple of weeks.
Best regards,
Mark Ouellet.
--- Squish v1.01
* Origin: The Sequel to Cramer VS Cramer: TPCramer VS UUencode ;-) (1:240/1.4)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:08:54
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 23:30:00
From Trevor Carlsen
To Niclas Pettersson
Subject Findfirst - SearchRec Name
MC>function FILLSTR (N : byte; S : string) : string;
MC>begin
MC> FILLSTR := Copy(S+' ',1,N)
MC>end;
NP> ...this rutine is slow and cumbersom! And it does not
NP> suport long strings. Way not somthing like this?
NP> FUNCTION SpaceStr(Size : Byte) : String;
NP> VAR TempStr: String;
NP> BEGIN
NP> FillChar(TempStr,SizeOf(TempStr),' ');
NP> TempStr[0] := Char(Size);
NP> SpaceStr := TempStr;
NP> END;
Your suggestion is completely different in its result to what is being requested
What is being attempted is to pad a string to a certain length.
Perhaps...
function FillStr(size: byte; ch: char; S: string): string;
var
TempStr: string;
begin
if length(s) < size then begin
FillChar(TempStr[length(s)+1],size-length(s),32);
move(s[1],TempStr[1],length(s));
TempStr[0] := chr(size);
FillStr := TempStr;
end else
FillStr := S;
end;
TeeCee
--- TC-ED v2.01
* Origin: The Pilbara's Pascal Centre (+61 91 732930) (3:690/644)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:09:46
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-17-92 16:27:00
From Trevor Carlsen
To Jud Mccranie
Subject Rules
JM> Is there a rule against asking for help if you have an
JM> illegal copy of the software, or a rule against helping
JM> someone who obviously has an illegal copy? If not, do you
JM> think it would be a good idea to have one?
Generally speaking, you will rarely, if ever, see long term seasoned contributor
assisting those asking questions when they "obviously" do not have a legal
copy of the compiler. I think that the guidelines laid down - "Do not use
the echo as a replacement for the manual you should have" is sufficient.
Unless the message openly admits pirating (stealing software) - such as that
by the idiot who claimed that he considered entitled to pirate because he
paid $7000 to some school or other - it is usually sufficient to just ignore
the message or point to the page number in the manual where the information
they seek is found.
TeeCee
--- TC-ED v2.01
* Origin: The Pilbara's Pascal Centre (+61 91 732930) (3:690/644)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:09:46
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-17-92 16:42:00
From Trevor Carlsen
To Jud Mccranie
Subject BP 7 difference
JM> The behavior of RANDOM (with RandSeed set) is different in
JM> BP7 (and presumably TP7) from that in TP 5.5. (I don't know
JM> how TP 6 compares since I burned it off my disk).
JM> RandSeed := 123;
JM> for i := 1 to 8 do writeln( random( 1000));
JM> TP 5.5: 343 282 986 996 781 855 343 32
JM> BP 7.0: 859 80 869 854 317 257 20 46
JM> ...both are consistant, but they are different sequences.
JM> This can have some dire consequences. ...
It certainly could if you did not know about it and unfortunately I can find
no reference to the changes in the documentation. (Richard Nelson?)
Here is a fix (supplied to me via Netmail courtesy Joe Lamoine - thanks Joe).
>Quote........
I posted a message on Compuserve last nite and got the following
unit in a response. It seems to work fine!
{ * Turbo Pascal Runtime Library Version 6.0 * ;
* Random Number Generator * ;
* * ;
* Copyright (C) 1988,92 Borland International * }
unit TP6Rand;
interface
function Random(Max: Integer): Integer;
implementation
const
{ Scaling constant}
ConstM31 = Longint(-31);
{ Multiplication factor}
Factor: Word = $8405;
function NextRand: Longint; assembler;
{ Compute next random number
New := 8088405H * Old + 1
Out DX:AX = Next random number
}
asm
MOV AX,RandSeed.Word[0]
MOV BX,RandSeed.Word[2]
MOV CX,AX
MUL Factor.Word[0] { New = Old.w0 * 8405H }
SHL CX,1 { New.w2 += Old.w0 * 808H }
SHL CX,1
SHL CX,1
ADD CH,CL
ADD DX,CX
ADD DX,BX { New.w2 += Old.w2 * 8405H }
SHL BX,1
SHL BX,1
ADD DX,BX
ADD DH,BL
MOV CL,5
SHL BX,CL
ADD DH,BL
ADD AX,1 { New += 1 }
ADC DX,0
MOV RandSeed.Word[0],AX
MOV RandSeed.Word[2],DX
end;
function Random(Max: Integer): Integer; assembler;
asm
CALL NextRand
XOR AX,AX
MOV BX,Max.Word[0]
OR BX,BX
JE @@1
XCHG AX,DX
DIV BX
XCHG AX,DX
@@1:
end;
end.
>End of quote.
TeeCee
--- TC-ED v2.01
* Origin: The Pilbara's Pascal Centre (+61 91 732930) (3:690/644)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:09:46
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-17-92 16:53:00
From Trevor Carlsen
To Nathan Wambach
Subject exe's
NW> OK, a program needs to know what it's own EXE file is
NW> called, how do you do it? What I wanted to use this for is
NW> linking files into the EXE... You have to know the EXE file
NW> name and who knows, maybe someone renamed it???
If DosVersion >= 3 then ProgramName := ParamStr(0)
TeeCee
--- TC-ED v2.01
* Origin: The Pilbara's Pascal Centre (+61 91 732930) (3:690/644)
* Tossed by SFToss/286 v1.02a on 92/12/18 09:09:46
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 07:59:00
From Dj Murdoch
To Mark Ouellet
Subject Re: borland pascal 7.0
MO> I must admit I have tried loading TD386.exe but I haven't tried
MO> installing TD386.sys BEFORE Qemm in my config. Maybe that could work.
I don't think TD386.EXE will work with QEMM no matter what you do, but unless
they've changed something, TD386.SYS will. You load it to get access to the
hardware debugging features of the 386 chip. They should be available in
TD and in TD286.
--- Msg V3.2
* Origin: Murdoch's_Point - - (1:249/99.5)
* Tossed by SFToss/286 v1.02a on 92/12/18 19:27:14
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 08:01:00
From Dj Murdoch
To Tom Lawrence
Subject Re: Speed wanted !
TL> Well, not quite. Borland's move uses movsb, and for
TL> some reason (probably a good one I'm not aware of), they
TL> point ds:si to the source variable, advance ds:si to the
TL> END of the source, then copy backwards (via std instead of
TL> cld).
That may be to handle overlapping source and destination blocks. If you have
var
bigarray : array [1..10000] of integer;
begin
move(bigarray[1],bigarray[2],9999*sizeof(integer));
end;
you'd better do the move in the backwards direction. The TP RTL checks the
addresses before it decides whether to go forwards or backwards.
--- Msg V3.2
* Origin: Murdoch's_Point - - (1:249/99.5)
* Tossed by SFToss/286 v1.02a on 92/12/18 19:27:14
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 08:10:00
From Dj Murdoch
To Mark Ouellet
Subject Re: Pointers
MO> I hit a wall while trying to fiddle with redirection and such stuff
MO> in BP 7.0.
MO> I basically declared a pointer to an Array [0..0] of <some type>
MO> When I tried assigning it a value:
MO> type
MO> FHT = Array[0..0] of byte;
Change this declaration to Array[0..65520] of byte, and your trouble should
go away.
--- Msg V3.2
* Origin: Murdoch's_Point - - (1:249/99.5)
* Tossed by SFToss/286 v1.02a on 92/12/18 19:27:14
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-16-92 08:12:00
From Dj Murdoch
To Craig Randle
Subject Re: Linking OBJ's
CR> Lately I've been trying to link an OBJ from a basic
CR> program into Pascal (v. 6.0), but I haven't been having
CR> much success. I know that the procedures are available
CR> for calling, but I keep coming up with the error undefined
CR> external. I am linking the OBJ through the { $L
CR> WHATEVER.OBJ } command. Does basic create OBJ that are
CR> uncompatible with TP or what?
It's not at all easy (and sometimes impossible) to link .OBJ files from another
high level language into TP. The TP "smart" linker doesn't understand .LIB
files, so the .OBJ file has to be completely self-contained, or only contain
references to externals that you're supplying from other .OBJ files or from
TP. If you have TDUMP (comes with TP 6 Professional or BP 7) you can see
a list of all the externals in the .OBJ file by dumping it, but you may not
be able to figure out what they all do in order to supply them.
--- Msg V3.2
* Origin: Murdoch's_Point - - (1:249/99.5)
* Tossed by SFToss/286 v1.02a on 92/12/18 19:27:14
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-18-92 13:12:00
From Trevor Carlsen
To Daniel Shapiro
Subject Turbo Pascal 6
> How about buying the real disks!!
DS> I don't have $100...All of us aren't well off..
So your practice is that if you don't have the money to buy something you
go out and steal it? An interesting concept; do you also apply it to cars
etc, or are you only a specialist in stealing software?
Moderator.
--- TC-ED v2.01
* Origin: The Pilbara's Pascal Centre (+61 91 732930) (3:690/644)
* Tossed by SFToss/286 v1.02a on 92/12/19 09:47:38
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-18-92 13:14:00
From Trevor Carlsen
To Vince Laurent
Subject Re: SB
VL> What about posting in multiple meessages? I thought
VL> the limit for FidoNet was 99 lines?
I am quite sure that some Reader authors would like you to believe that. FidoNet
messages have no official limit. This echo has a 16K message limit with as
many lines as you can fit within that limit.
VL> ...(I would still like to get my hands on that code).
As I said... it was almost certainly in the packet you received but your reader
truncated it. If it is important to you, arrange a private exchange via netmail
or snailmail.
Moderator
--- TC-ED v2.01
* Origin: The Pilbara's Pascal Centre (+61 91 732930) (3:690/644)
* Tossed by SFToss/286 v1.02a on 92/12/19 09:47:38
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-17-92 08:25:00
From Dj Murdoch
To John Lester
Subject Re: Dumping Arrays...
JL> I'm working on a program that uses quite a few arrays. Is there
JL> anyway in TP 5.5 to dump them out of memory so that I can load
JL> the next set of arrays for processing?
Sure, store them on the heap.
--- Msg V3.2
* Origin: Murdoch's_Point - - (1:249/99.5)
* Tossed by SFToss/286 v1.02a on 92/12/19 09:48:14
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-17-92 08:28:00
From Dj Murdoch
To Joe Jared
Subject Re: HUFFMAN.PAS
JJ> ote the size difference between your program and the 2
JJ> below. Either you're not doing a full huffman
JJ> compression, or there's a bug somewhere (InsertNode?). It
JJ> also explains why your version of compress is faster than
JJ> my assembler version. I wouldn't be at all surprised if
JJ> your code doubles in time if you implement full huffman
JJ> compression or fix the bug.
What do you mean by "full Huffman compression"? I'm compressing bytes according
to a simple version of the algorithm that I found in a book by McEliece (?).
It just generates a byte count table, then pools the pair of least frequent
entries and inserts the new node in the appropriate place.
There are a couple of places where it makes an arbitrary decision about what
to do in the case of a tie, but I can't see many other choices. Can you describ
the algorithm you're using, or point me to a reference to it?
--- Msg V3.2
* Origin: Murdoch's_Point - - (1:249/99.5)
* Tossed by SFToss/286 v1.02a on 92/12/19 09:48:14
▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄▄
Conference 4
Date 12-17-92 08:39:00
From Dj Murdoch
To Tom Lawrence
Subject Re: Help!
TL> Now, not to start up this war again (he he), but could you explain WHY
TL> using a GOTO is considered DIRTY programming, but using a
TL> break is not?? The
TL> machine code generated is IDENTICAL in both cases. Just curious :)
The machine code has nothing to do with it. It's the meaning of the program
that's different.
If you use break, you know that when you're on the statement following the
loop, the only way to get there is to have finished the loop. "Finished"
now means that you've hit the end of the loop or hit break; clearly not using
break is preferable to using it, but there are cases where you don't want
to go all the way through the loop.
If you use a goto, you'll put a label on the statement following the loop.
When you look at that label, you have no idea at all how you got there.
You have to read the whole program in order to find the Goto(s) that took
you there. (In TP, you only have to read the whole procedure/function.)
The way to make a program reliable is to always be absolutely sure of what
state it's in at any time. It's possible to use Goto in such a way that this
is easy, but there's no need to: Pascal (especially TP 7) provides lots of
choices of control structures that allow you to specify more precisely what
you want.
--- Msg V3.2
* Origin: Murdoch's_Point - - (1:249/99.5)
* Tossed by SFToss/286 v1.02a on 92/12/19 09:48:14