Celestin Apprentice 5/Apprentice-Release5.iso/Information/THINK C Digest/1993/93-07

Text File | 1995-12-31 | 116KB | 3,506 lines

Path: ucivax!gateway
From: [e-mail] (Gary Au)
Subject: flight simulator in THINK C
Message-ID: <[e-mail]>
X-Mailer: ELM [version 2.3 PL11]
Newsgroups: fa.think-c
Lines: 14
Date: 2 Jul 93 07:16:00 GMT

Hi!

I am writing a freeware air to air combat/flight simulator using THINK C. I've
seen lots of IBM air combat games, and was thinking of doing something
similar for the mac - something like Falcon, but for low level nape of the
earth flying... lots of mountains flying past etc. Does anyone know the
secrets of flight sim algorithms? An idea was to store the ground as
a 2D array, and do a 3D perspective job on the thing. The drawback of this
technique is that the graphics take too long to draw. I mean, calling
Lineto all the time is CPU intensive. How about storing the current position
as a point, and drawing the graphics depending on where that point is?

gary











Path: ucivax!gateway
From: [e-mail] (Dana S Emery)
Subject: Re: flight simulator in THINK C
Message-ID: <[e-mail]>
In-Reply-To: Your message <[e-mail]> of 2
 Jul 93 07:16:00 GMT
Content-Type: TEXT/plain; charset=US-ASCII
Newsgroups: fa.think-c
Lines: 13
Date: 2 Jul 93 16:38:18 GMT

off screen pixmaps is the only way to go, there you know you are in RAM, and can
use non-quickdraw drawing techniques with impunity.

The subsequent screen update is done efficiently and legally.

Develop magazine and Mactutor (now MacTech) have had several articles on this
and on GWorld manipulation, you need familiarity with both.

dont have indexes on hand, but develop is cheap enough to get on CD rom, talk to
APDA.
--
dana s emery <[e-mail]>











Path: ucivax!gateway
From: [e-mail] (Frank Deming x6088)
Subject: Re: flight simulator in THINK C
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 4
Date: 2 Jul 93 16:54:32 GMT

There is a book out on writing a flight simulator and using c++. The name
escapes me at the moment

Frank Deming










Path: ucivax!gateway
From: [e-mail] (Greg Ferguson)
Subject: Re: flight simulator in THINK C
Message-ID: <[e-mail]>
Content-Type: text/plain; charset="us-ascii"
Mime-Version: 1.0
Newsgroups: fa.think-c
Lines: 36
Date: 2 Jul 93 18:00:43 GMT

At  7:16 AM 7/2/93 +0000, Gary Au wrote:
>I am writing a freeware air to air combat/flight simulator using THINK C. I've
>seen lots of IBM air combat games, and was thinking of doing something
>similar for the mac - something like Falcon, but for low level nape of the
>earth flying... lots of mountains flying past etc. Does anyone know the
>secrets of flight sim algorithms? An idea was to store the ground as
>a 2D array, and do a 3D perspective job on the thing. The drawback of this
>technique is that the graphics take too long to draw. I mean, calling
>Lineto all the time is CPU intensive. How about storing the current position
>as a point, and drawing the graphics depending on where that point is?

The guy who wrote Microsoft's Flight Simulator wrote a book about graphics
on the IBM. I think he discusses the techniques needed for that style of
program. Unfortunately, I'm at work and the book is at home, and I haven't
looked at it in so long I can't remember what it's called. I'll try to
remember to look tonight and follow up later.

Greg


-=-=-=-=-=-=-
Greg Ferguson                           email:  [e-mail]
Programmer/Engineer                     voice:  602.952.3623
Desktop Networking Systems
World Marketing
Motorola, SPS

*************************************************************
Don't worry, though, it will be just as easy to spot
Pentium-equipped PCs as it is to spot PCs now. Almost all of
them come with this useful little warning required by the
Truth In Advertising Act, saying "Intel Inside."
        - Pythaeus in TidBits
*************************************************************












Path: ucivax!gateway
From: [e-mail] ("A. Scottedward Hodel")
Subject: f2c and things you can't find.
Full-Name: A. Scottedward Hodel
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 9
Date: 2 Jul 93 18:24:06 GMT


I'm trying to port a small C++ numerical package I wrote for a Unix
dumb terminal in C++ onto my Mac.  Much of the code is Fortran
converted to C++ with f2c (recently posted to info-mac).  The problem
is that I'm regularly encountering undefined functions/incorrect
numbers of arguments, etc., even in the simple f2c example, when
using Symatec C++.  Is there some relatively painless procedure to
identify which headers/libraries one should use when these errors
arise?










Path: ucivax!gateway
From: [e-mail] (Jim Lynch)
Subject: Re:  flight simulator in THINK C
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 14
Date: 2 Jul 93 22:20:30 GMT

Math-wise, you need calculus (to find rates of change) and physics.

Assume you have a position function based on time, gravity, weight, lift, wind
and thrust (together with direction of lift, of thrust and of wind).

If you took calculus you know how to differentiate that position function
which gives you speed (and differentiating that gives you acceleration.)

Then all you would need are the specifics of the plane.

For more realism, you could map wind patterns such as those that surround
mountains or wind events such as thermals, wind shears and tornado/hurricanes.

Of course, speed suffers as you take more things into account.










Path: ucivax!gateway
From: [e-mail] ("Mark D. Nagel")
Subject: Re: flight simulator in THINK C
Message-ID: <[e-mail]>
X-Mailer: uAccess - Macintosh Release: 1.6v1
Newsgroups: fa.think-c
Reply-To: [e-mail]
Lines: 15
Date: 3 Jul 93 15:36:53 GMT


In Regards to your letter <[e-mail]>:
> There is a book out on writing a flight simulator and using c++. The name
> escapes me at the moment

You are probably thinking of Flights of Fantasy by Christopher Lampton (ISBN
1-878739-18-2).

It's a cool book, but its details focus on PC graphics, and many have noted
various flaws in the code.  However, it goes into 3D polygon rendering, hidden
surface removal, and other stuff you need for flight simulators.  It comes
with a disk containing source and executables, but the executables are for
the PC, of course.

Mark










Path: ucivax!gateway
From: [e-mail] (pfterry)
Subject: CMaster 1.2 -> 1.2.3 updater
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 23
Date: 3 Jul 93 16:50:00 GMT



I have uploaded the CMaster 1.2.3 updater to the ThinkC and TCL ftp sites and
sent it to MacGifts. It should appear on the other popular ftp sites after the
holiday. Here's the description uploaded with the updater.

This updater converts CMaster 1.2.1 or 1.2.2 to version 1.2.3. This  newest
release contains a fix to the option-dbl-click problem and scrolling to
preceding comments should now work correctly. Key equivalents for THINK C 6
changed, so some of our key binding routines changed to reflect these. A couple
 other minor fixes including the v scroll bar pixel that was erased sometimes.

Enjoy.

pf

+++++++++++++++++++++++++++++++++++++++++++++
Fred Terry, Kansas Geological Survey
1930 Constant Ave., Lawrence, KS  66047
[e-mail]
+++++++++++++++++++++++++++++++++++++++++++++

I've tried to relax, but--I dunno--I seem to be more comfortable tense.










Path: ucivax!gateway
From: [e-mail] (pfterry)
Subject: Another ftp site for the CMaster updater
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 17
Date: 3 Jul 93 16:56:37 GMT



Just in case someone is actually trying to get something done this week and
wants/needs the updater, I've placed it on syrinx.kgs.ukans.edu. You'll find it
in topsoft/other.stuff/.

Enjoy.

pf

+++++++++++++++++++++++++++++++++++++++++++++
Fred Terry, Kansas Geological Survey
1930 Constant Ave., Lawrence, KS  66047
[e-mail]
+++++++++++++++++++++++++++++++++++++++++++++

I've tried to relax, but--I dunno--I seem to be more comfortable tense.










Path: ucivax!gateway
From: [e-mail] ("Mark D. Nagel")
Subject: Re: flight simulator in THINK C
Message-ID: <[e-mail]>
X-Mailer: uAccess - Macintosh Release: 1.6v1
Newsgroups: fa.think-c
Reply-To: [e-mail]
Lines: 16
Date: 3 Jul 93 17:27:48 GMT


In Regards to Dana S Emery's letter <[e-mail]>:
> off screen pixmaps is the only way to go, there you know you are in RAM, and
> can use non-quickdraw drawing techniques with impunity.
>
> The subsequent screen update is done efficiently and legally.

This is true to an extent, but CopyBits is often too slow for games
where you have to get 15 or more frames per second (ignoring those
games like Maelstrom and Space Madness which are designed to move as
little of the screen around as possible per frame).

Most games have to go direct-to-screen and build in compatibility modes
to allow future machines to use Quickdraw in case Apple breaks the current
methods for drawing directly to the screen (which they promise will happen
when the PowerPC comes out).










Path: ucivax!gateway
From: [e-mail] (Christopher Evans)
Subject: PD regexp or grep source
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 17
Date: 5 Jul 93 07:16:22 GMT


    I need to write a text find function for a text editor and would like
to include a Regular Expression option.  Can anyone point me towards some
source code that will make my job easier?

    Thanks!

<==================================Q======================================>
  Christopher Evans                |   Internet: [e-mail]
  Director, Software Development   |   Phone: [phone]
  Natural Intelligence, Inc.       |   FAX:   [phone]
  2067 Massachusetts Avenue        |   AppleLink: NATURAL
  Cambridge MA  02140              |   or [e-mail]@INTERNET#

    If God is perfect, why did He create discontinuous functions?












Path: ucivax!gateway
From: [e-mail] (Rodolfo Cardarel)
Subject: Bandwith waste
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 9
Date: 5 Jul 93 12:23:27 GMT

Maybe it's just a stupid suggestion, but, as lots of people keep on
misunderstanding the correct usage of think-c and think-c-request (I did the
same mistake myself...) why not changing the two account names to something
more understandable, as, for example, administration-think-c and q&a-think-c?
And, by the way, could you give me the correct location of the f2c s/w, it
should be on info-mac of sumex-aim, but I'm not sure.
Thank you

Rodolfo Cardarelli [e-mail]










Path: ucivax!gateway
From: [e-mail] (Jean Berthomieu)
Subject: Re: Bandwith waste
Message-ID: <[e-mail]>
In-Reply-To: <[e-mail]> from "Rodolfo Cardarel" at Jul 5, 93 12:23:27 pm
X-Mailer: ELM [version 2.4 PL22]
Content-Type: text
Content-Length: 1087
Newsgroups: fa.think-c
Lines: 23
Date: 5 Jul 93 15:43:25 GMT

>
> Maybe it's just a stupid suggestion, but, as lots of people keep on
> misunderstanding the correct usage of think-c and think-c-request (I did the
> same mistake myself...) why not changing the two account names to something
> more understandable, as, for example, administration-think-c and q&a-think-c?
> And, by the way, could you give me the correct location of the f2c s/w, it
> should be on info-mac of sumex-aim, but I'm not sure.
> Thank you
>
> Rodolfo Cardarelli [e-mail]
>
R1) There would be people who would misunderstand the correct usage
    of administration-think-c and q&a-think-c anyway!
    and since it is a standard on Internet to call xxx-request the xxx list
    manager address, why change it now?

R2) The original repository of f2c is research.att.com in /netlib/f2c
    There you will find the latest release.
    At prep.ai.mit.edu you may find one (or two) previous release which may
    be more fitted to your Unix(?) system in /pub/gnu.
    There has been a recent posting of f2c - for THINK_C - at:
    sumex-aim.stanford.edu.











Path: ucivax!gateway
From: [e-mail] (Dana S Emery)
Subject: Re: Bandwith waste
Message-ID: <[e-mail]>
In-Reply-To: Your message <[e-mail]> of 5 Jul 93
 12:23:27 GMT
Content-Type: TEXT/plain; charset=US-ASCII
Newsgroups: fa.think-c
Lines: 19
Date: 5 Jul 93 18:54:41 GMT

>   Maybe it's just a stupid suggestion, but, as lots of
>   people keep on misunderstanding the correct usage of
>   think-c and think-c-request

mumble-request is the traditional way to name the admin address of a list named
mumble.

thats the standard folks, good or bad.  Several hundred lists use it, any
attempt to change it now would likely fail, and would certainly cause lots of
confusion.

Interestingly, all the lists I am a member of (including Think-c) issue a
welcomeing message to newly subscribed members which explains this.

I dont know about you, but I take the time to both read and archive the welcomes
I get.
--
dana s emery <[e-mail]>











Path: ucivax!gateway
From: [e-mail]
Subject: Help reading data from file
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 27
X-Mts: smtp
Date: 5 Jul 93 19:27:38 GMT

    I am trying to write a code in Think C (Actually Think
Lightspeed C 2.15...) that reads a file character-by-character
(actually, it should be alphanumeric, but...) and process it according
to the type of character.  The algorythm would be something like:

Open infile and outfile

While you did not reach EOF:
{
    fget (infile, char) (I think)

        case char
        {

            uppercase: do something
            lowercase: do something else
            number: handle accordingly

        }

    Write modified stuff in outfile
}
close files


How can I do such think under Think C?  Can I use fget/fput as I seem
fit?  If not, is there some kind of Mac-exclusive function I should use?










Path: ucivax!gateway
From: [e-mail] (Stefan Arentz)
Subject: (Q) Beta-Testers Wanted
X-Sender: [e-mail] (Unverified)
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 49
Date: 5 Jul 93 21:22:15 GMT

Hello World,

I'm looking for beta-testers for a library I am writing. It's a
replacement for the Dialog Manager, because the Dialog Manager
has some shortcomings.


Here are the features of the 'XDialogManager' library:

 - More kinds of item types: OKButton, CancelButton, Button,
   CheckBox, RadioButton, List, Border, EditText, StaticText,
   UserItem, Icon, Picture, PopupMenu.

 - Button and text items can use custom fonts (specified in the
   XDTL resource).

 - Radio buttons have a groups ID number. The library
   automatically updates radio button groups.

 - Buttons can have a command-key equivalent, like BBEdit does.

 - All items can be visible/invisible and enabled/disabled. Text
   items are dimmed.

 - Better userProc for UserItems (draw & click messages).

 - Compatibility with System Seven and Six.

 - Interface for C and Pascal.

 - Libraries for MPW C/Pascal and THINK C/Pascal compilers.


And I'm working on:

 - ResEdit Editor for 'XDTL' resources.


If you think this library is useful for you, then please respond,
to this message and I will send you the latest version. I would
like to hear from you what you think of the API, if you found
bugs and if you miss a feature in this library.

-- Stefan


----------------------------------------------------------------------------
Stefan Arentz          -  Software Constructor  -          [e-mail]











Path: ucivax!gateway
From: [e-mail] (Stefan Arentz)
Subject: (Q) Popup Menus in Windows
X-Sender: [e-mail] (Unverified)
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 19
Date: 5 Jul 93 21:22:19 GMT

Hello World,

I'm having trouble with popup menus in a window. I install the popup
menu with _NewControl. If a mouseDown event in the window occurs and
if _FindControl says it's my control, then I call _TrackControl. The
strange thing is that _TrackControl works with the Buttons,
RadioButtons and CheckBoxes in the window, but not with the popup
menu. The popup menu's title is tracked (and highlighted), but the
menu never pops up. The 'MENU' resource isn't missing, because the
first item of menu is set to the first item in the 'MENU' resource.

Any Ideas? What am I doing wrong?

-- Stefan


----------------------------------------------------------------------------
Stefan Arentz          -  Software Constructor  -          [e-mail]











Path: ucivax!gateway
From: [e-mail] (Johan Strandberg)
Subject: Re: Help reading data from file
X-Sender: [e-mail] (Unverified)
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 174
Date: 6 Jul 93 02:13:16 GMT

At  7:27 PM 7/5/93 +0000, [e-mail] wrote:
>        I am trying to write a code in Think C (Actually Think
>Lightspeed C 2.15...)

You really shouldn't use that old a version of Think C since it has serious
compatibility problems with recent 680x0 processors (at least the 040 and
030) that would make the code unreliable on any recent Macintosh.
(Basically there is some missing flushes of the instruction cache.)  I
forget at what version it was fixed -- somebody out there no doubt
remembers.  I understand that there might be economical reasons not to
upgrade, but you should remember this warning when you run into strange
problems.

The code below was tested under THINK C 6.0.  I don't remember all the
restrictions to version 2.15, so there might be some problems compiling the
code below straight as it is under your compiler.  What I have written is
pretty much Standard C (it also compiles under C++).


>that reads a file character-by-character
>(actually, it should be alphanumeric, but...) and process it according
>to the type of character.  The algorythm would be something like:
>
>Open infile and outfile
>
>While you did not reach EOF:
>{
>        fget (infile, char) (I think)
>
>                case char
>                {
>
>                        uppercase: do something
>                        lowercase: do something else
>                        number: handle accordingly
>
>                }
>
>        Write modified stuff in outfile
>}
>close files
>
>
>How can I do such think under Think C?  Can I use fget/fput as I seem
>fit?  If not, is there some kind of Mac-exclusive function I should use?

Here is some code that will do what you want.  Replace the code inside the
while statement with whatever you wish.

If the following is in the file example.in:

        ABCDEFGHIJKlmnopqrstuvxyz0123456789+-
        Four scores and 7 years ago...

The the file example.out will become:

        XXXXXXXXXXXxxxxxxxxxxxxxx9999999999..
        Xxxx xxxxxx xxx 9 xxxxx xxx...


Good Luck!


--johan




/*======================================================================
    Example of file I/O in C by Johan Strandberg, 1993
    This code is in the public domain.  Share and Enjoy!
======================================================================*/
#include <stdio.h>
#include <ctype.h>
#include <errno.h>



/*======================================================================
|           transformFile()
|-----------------------------------------------------------------------
|   Open the inFile and outfile.  Read each character of inFile, and
|   for each character do some processing.
|
|   Please note that the "functions" isupper(), islower(), etc are
|   *very* fast since they are normally implemented as macros that do
|   table lookup.  Never write:
|
|           if ('A' <= ch && 'Z' >= ch)
|
|   instead use:
|
|           if (isupper(ch))
|
|   This generates faster code (and more correct, the above doesn't
|   always work with non-ASCII character sets).
|
|
|   Takes:
|       inFile      Pointer to the name of the input file.
|       outFile     Pointer to the name of the output file.  This file
|                   will be replaced if it is already there and created
|                   if it is not.  There is no explicit code to set
|                   the Type and Creator of the file, so the actual
|                   Type/Creator will depend on the I/O library used
|                   by the compiler.
|
|   Returns:    0   Success.
|               !0  The error code.
|---------------------------------------------------------------------*/
int transformFile(const char *inFile, const char *outFile)
{
    FILE    *in;        /* File pointer to the input file */
    FILE    *out;       /* File pointer to the output file */
    int     err;        /* To store temporary error values */
    int     ch;         /* Make sure it's an int! */


    errno = 0;
    in = fopen(inFile, "r");
    if (errno) {
        return errno;
    }

    errno = 0;
    out = fopen(outFile, "w");
    if (errno) {
        err = errno; /* save errno since fclose will scramble it */
        fclose(in);
        return err;
    }

    while ((ch = fgetc(in)) != EOF) {
        /* This section would contain whatever transformations you want */
        if (isupper(ch)) {
            fputc('X', out);    /* Replace all upper case characters with 'X' */
        } else if (islower(ch)) {
            fputc('x', out);    /* Replace all lower case characters with 'x' */
        } else if (isdigit(ch)) {
            fputc('9', out);    /* Replace all digits with '9' */
        } else if (isspace(ch)) {
            fputc(ch, out);     /* Preserve all whitespace */
        } else {
            fputc('.', out);    /* Replace all other characters with '.' */
        }
    }

    /*  Intentionally ignore all error messages during fclose.
        There isn't much else to do if it fails anyway...
    */
    fclose(in);
    fclose(out);

    return 0;
}


/*======================================================================
|           main()
|-----------------------------------------------------------------------
|   This is just a simple main() to test the transformFile code.  Since
|   this is just an example, we don't have code for fetching the
|   file names here.  Instead we just use the fixed names of example.in
|   and example.out when we call the transformFile example.
|
|   Returns:    0   - Always succeeds
|---------------------------------------------------------------------*/
int main(void)
{
    transformFile("example.in", "example.out");

    return 0;
}











Path: ucivax!gateway
From: [e-mail] (Johan Strandberg)
Subject: Re: PD regexp or grep source
X-Sender: [e-mail] (Unverified)
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 32
Date: 6 Jul 93 02:13:22 GMT

At  7:16 AM 7/5/93 +0000, Christopher Evans wrote:
>        I need to write a text find function for a text editor and would like
>to include a Regular Expression option.  Can anyone point me towards some
>source code that will make my job easier?
>
>        Thanks!


You should look into Henry Spencer's regexp library.  It is the industry
standard for all practical purposes and it's free and
"nearly-public-domain".  The README file states in part:

        Permission is granted to anyone to use this software for any
        purpose on any computer system, and to redistribute it freely,
        subject to the following restrictions:

        1. The author is not responsible for the consequences of use
           of this software, no matter how awful, even if they arise
           from defects in it.

        2. The origin of this software must not be misrepresented,
           either by explicit claim or by omission.

        3. Altered versions must be plainly marked as such, and must
           not be misrepresented as being the original software.

It's available from *many* archive sites on the net (use archie to get a
fresh list), just look for regexp.


--johan











Path: ucivax!gateway
From: [e-mail] (David Sugar)
Subject: Questions???
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 25
Date: 6 Jul 93 13:46:17 GMT


  I am currently working on a program and can't seem to figure out how to
do some things that I want to do, they don't seem like they would be to
hard, but I can't find them mentioned.

  The first should be real easy.  I need to build a menu that looks just
like the finder's "label" menu.  I want to give the user a choice of a
label, and using the same menu (same text, same colors etc) seems like
the best way to do it, but I can't find the label menu stored anywhere.
Is there a way to access this info from a program??

  Second question, a little bit harder.  I want a way to open a finder
window from this program.  Like if I have a directory ID of some directory
in this application I would like to be able to automaticly open this
window in the finder for the user.  (I know that I should only do this
after asking the user if I should do it or not and the user interface
stuff)..


  Thanks for any answers
  Dave Sugar
  [e-mail]
  [e-mail]












Path: ucivax!gateway
From: [e-mail] ("Paul S. Michelsen")
Subject: Re: (Q) Beta-Testers Wanted
Message-ID: <[e-mail]>
In-Reply-To: <[e-mail]> (message from Stefan Arentz on 5 Jul 93 21:22:15 GMT)
Newsgroups: fa.think-c
Lines: 34
Date: 6 Jul 93 16:39:35 GMT

  >Hello World,
  >
  >I'm looking for beta-testers for a library I am writing. It's a
  >replacement for the Dialog Manager, because the Dialog Manager
  >has some shortcomings.
  >
  >Here are the features of the 'XDialogManager' library:

   [list of goodies]

  >And I'm working on:
  >
  > - ResEdit Editor for 'XDTL' resources.
  >
  >If you think this library is useful for you, then please respond,
  >to this message and I will send you the latest version. I would
  >like to hear from you what you think of the API, if you found
  >bugs and if you miss a feature in this library.
  >
  >-- Stefan
  >
  >
  >----------------------------------------------------------------------------
  >Stefan Arentz          -  Software Constructor  -          [e-mail]

I'm interested in having a look at it.  Could you send me a copy?
Thanks.  (I assume it comes with full source code and documentation.)

Thanks.

=|<>|=  Paul S. Michelsen                  Eventually, computers will be
=|  |=  [e-mail]              infinitely fast, impossibly small,
=|  |=  ext x23224, rm M2-160                       and free.
=|__|=  Amdahl, Sunnyvale, California










Path: ucivax!gateway
From: [e-mail] (Jeremy John Ahouse)
Subject: HT like string manipulations
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 31
Date: 6 Jul 93 22:38:54 GMT

Does anyone know if there is a library for C or C++ that implements
container manipulations like HyperTalk.  Manipulating strings is so easy in
HyperTalk it would be great to do this in C.  I may have to write it myself
but if someone already has... Of course it doesn't have to be == to HT and
rich set of string manipulations would be nice... ideas??

        Thanks,

        - Jeremy

        :::::::::::::::::::::::::::::::::::::::::::
        Jeremy John Ahouse
        Biology Dept and Center for Complex Systems
        Brandeis University
        Waltham, MA 02254-9110

        [phone]
        email: [e-mail]
        Mail from Mac by Eudora 1.3.1

        "Si un hombre nunca se contradice, sera porque nunca dice nada"
                - Miguel de Unamuno

RIPEM accepted.
PublicKeyInfo:
 MH8wCgYEVQgBAQICAzYDcQAwbgJnNZ6A64065Jt9uKprRlOWtbi1N7C82eQcmse3
 ahK9j3Y86tayGz9iqg0OIBhQFe7qRqxW+j6MM81sxlxDf23x1ecvyva9uhaLPLJR
 2Kttl56iG3u11WWcyf7SeW3dW/JW0DQ5gnTqUwIDAQAB
MD5OfPublicKey: 0E6662766C4D0BDAB766A6B4DA80242A












Path: ucivax!gateway
From: [e-mail]
Subject: Library wanted...
Message-ID: <[e-mail]>
Content-transfer-encoding: 7BIT
X-VMS-Cc: lafarciola
Newsgroups: fa.think-c
X-VMS-To: in%"[e-mail]"
Lines: 9
Date: 7 Jul 93 08:57:44 GMT

Hello!
Where can I find a statistical library?
Is there an archive where I can ftp it?

Thanks for any help,
        -Massimo-

E-mail: [e-mail]











Path: ucivax!gateway
From: [e-mail] (Big Nose)
Subject: THINK Reference
Via: UK.AC.AFRC.IAPE;  8 JUL 93 15:48:52 BST
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 10
Date: 8 Jul 93 14:46:55 GMT


I know this probably isn't strictly a proper question, but would anyone
care to tell me the current US Street price for THINK Reference. It's a
possibility that I may get a friend to buy it for me when he's over, if
it's much cheaper that is.

Thanks in advance,

Andy Law
( LAWA @ IAPE.AFRC.AC.UK                       Big Nose in Edinburgh )










Path: ucivax!gateway
From: [e-mail] (Chris Dixson)
Subject: Readings Files
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 18
Date: 8 Jul 93 18:04:10 GMT


 I'm writing a program that converts rtf files.  So far this app allows
the user to select the file to convert, runs thru a simple conversion, and
then allows the user to save the converted file with a new name.  These
parts all work okay.  (For debugging purposes I've put up a console window
to display the text as it is read in and converted but the final app will
not display this text to the user).

 What I want to do next is implement a mappings file. This would compare
lines in the original file with this mapping file and substitute specific
rtf code with new rtf code. Any suggestions on the best way to implement
this? I'm having problems just reading in the mappings file. It's in the
same dir as the app itself but I'm having trouble using GetSFCurDir to get
the current dir and then building the FSSpec. I know StandardGetFile builds
the FSSpec but how do you open a specific file in a specific dir without
using StandardGetFile?  Any help would be greatly appreciated! :-)

 Thanks,        -Chris










Path: ucivax!gateway
From: [e-mail] (Kirk Crawford)
Subject: Think C 6.0
Message-ID: <[e-mail]>
Posted-Date: 7 Jul 93 16:00:57 U
Newsgroups: fa.think-c
Lines: 11
Date: 8 Jul 93 21:04:42 GMT

Subject:  Think C 6.0
Well, I have the new beast installed.  I like it overall, but I do have one
peeve right off the bat.

When I open a project file with it, it does not set the directory that standard
file uses to the directory with the project in it.  So, after opening a project
via an alias or something, to open a text file in the same folder I have to
navigate with STDFile again.  Think C 5 did not do this.  Any ideas?  Is an
update due out soon?












Path: ucivax!gateway
From: [e-mail]
Subject: Think 6.0 Extentions Technical info
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
X-Origin: The Portal System (TM)
Lines: 10
Date: 9 Jul 93 00:10:27 GMT


The new Think C Project Manager allows for extentions, like a C compiler, C++
compiler, etc. I have an stand alone assembler that I would like to put into
this enviroment. Were can I get technical documentation on how to write an
extention?

Aloha,
Steve Hales

HALESTORM, Incorporated










Path: ucivax!gateway
From: [e-mail] (Chuck Shavit)
Subject: SourceServer Bug
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 7
Date: 9 Jul 93 02:58:58 GMT

This may seem insignificant compared to the other SC++ bugs, but here it is
anyway.  The SourceServer produces some funny error messages at times and
refuse to do things like create a new ProjectDB, or check stuff in.  After
spending some time on it, I found that it happens when the pathname of the DB
file has a ' (quote) in it, e.g. if the pathname is Joe's:proj1:my_stuff.

Chuck Shavit










Path: ucivax!gateway
From: [e-mail] (Johan Strandberg)
Subject: Re: SourceServer Bug
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 36
Date: 9 Jul 93 04:28:25 GMT

>This may seem insignificant compared to the other SC++ bugs, but here it is
>anyway.  The SourceServer produces some funny error messages at times and
>refuse to do things like create a new ProjectDB, or check stuff in.  After
>spending some time on it, I found that it happens when the pathname of the DB
>file has a ' (quote) in it, e.g. if the pathname is Joe's:proj1:my_stuff.
>
>Chuck Shavit

You need to quote the name for source server if that's the case.  Try
something like

        Joe@'s:proj1:my_stuff

where @ is the option-d character.  I'm not sure of the exact rules, but
something along those lines might work.

In general the SourceServer documentation is lousy in my opinion, I have a
bunch of code that talks AppleEvents directly to it and there is *no* way
you can be expected to figure out what SourceServer needs in the event
strewam from the provided documentation.  For example, to send something
like:

        mountproject mydisk:myproject:

you need to send a list containing "mountproject" and "mydisk:myproject:"
as separate entries, but just not normal strings but null terminated
C-style strings.  If you know AppleEvents this is *weird*.  What's even
kinkier that even though the data stream is nicley tokenized -- as
described above -- SourceServer *still* needs quotes and stuff to handle
special characters.

Just a pain, that's all...


--johan











Path: ucivax!gateway
From: [e-mail] (Big Nose)
Subject: MenuBars/popupmenus and hook routines
Via: UK.AC.AFRC.IAPE;  9 JUL 93  9:09:27 BST
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 22
Date: 9 Jul 93 08:08:28 GMT


Dear All,

For various reasons, I need to use several popupmenus from a patch I am
writing. I would like them to behave like a normal menu bar, cf the
menus that Boomerang or Directory Assistance put into the SF Dialog. I
thought that I could use PopUpMenuSelect and use the MenuHook procedure
to determine when control from one menu should be aborted and passed on
to a different one i.e. when the user moves from one menu title in the bar
to the next, without releasing the mouse. Unfortunately, from what I can
determine, PopUpMenuSelect doesn't seem to call the MenuHook routine.

Anyone suggest how I can get round this. Is there a PopUpMenuHook routine
that is different from the standard MenuHook?, can I patch the Menu Manager
to draw a menu bar somewhere other than at the top of the screen? Does
patching SystemEvent (or similar) and posting fake mouseUp/mouseDown
combinations make the most sense?

Thanks,

Andy Law
( LAWA @ iape.afrc.ac.uk                        Big Nose in Edinburgh )










Path: ucivax!gateway
From: [e-mail] (Kevin Quick)
Subject: Appmaker Updater
Via: uk.ac.open; Fri, 9 Jul 1993 11:59:57 +0100
X-Sender: kc_quick@sunflower
Message-ID: <[e-mail]>
X-Mailer: Eudora 1.3.1
Newsgroups: fa.think-c
Lines: 25
Date: 9 Jul 93 11:02:01 GMT

Does anyone have the full Appmaker upgrade 1.5.2 > 1.5.4 for Symantec C
6.00. Someone did try to post it to Sumex but only included the
application, not the files that go with it.

If the update cost money can someone let me know the price.

As I've now got my SC++ upgrade I would like to continue to use this app as
well.

TIA

 ****************************************************************************
 *                                                                          *
 *  Kevin Quick                                                             *
 *  Computer Support Engineer          Telephone    [phone] (DDI)       *
 *  The Open University                Fax          [phone]             *
 *  Faculty of Technology              EMail        [e-mail]    *
 *  Venables Building                                                       *
 *  Walton Hall                                                             *
 *  MK7 6AA                                                                 *
 *  England                                                                 *
 *                                                                          *
 ***************************************************************************
*











Path: ucivax!gateway
From: [e-mail] (Phil Shapiro)
Subject: Re: Think 6.0 Extentions Technical info
X-Sender: [e-mail]
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 21
Date: 9 Jul 93 16:02:28 GMT

>The new Think C Project Manager allows for extentions, like a C compiler, C++
>compiler, etc. I have an stand alone assembler that I would like to put into
>this enviroment. Were can I get technical documentation on how to write an
>extention?

If you are interested in writing a THINK Translator (like a compiler or
assembler), you should send an email to the Product Manager, David Allcott,
at:  [e-mail]

You should also contact him if you're interested in writing an External
Editor that works with the THINK Project Manager, or if you're interested
in taking advantage of the built in editor hooks that the THINK editor
supports.

        -phil
--
   Phil Shapiro                                   Software Engineer
   Language Products Group                     Symantec Corporation
                 Internet: [e-mail]
                      "relatively inexperienced"











Path: ucivax!gateway
From: [e-mail] (Brian Williams)
Subject: Dagnabit_ResFiles
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 38
Date: 9 Jul 93 16:44:47 GMT


Hello all,

I am having a problem with resource files.  I am running myproj.pi
(TC 5.0), and would like to open temp.pi.rsrc to get a picture
and print it. Here is what I tried...

    Str255 resfile;
    int refNum, res_id, error;
    PicHandle thePicture;


    strcpy (resfile, "temp.pi.rsrc");
    CtoPstr (resfile);
    refNum = OpenResFile (resfile);
    if (refNum == -1)
    {
        errmessage :  cannot open file
    }

    UseResFile (refNum);
    error = ResError ();
    if (error != noErr)
    {
        errmessage :  resError = ...
    }
    thePicture = GetPicture (res_id);

    /* print the picture */

When I run, I do not receive any error messages, just a blank picture.
When I copied the picture from temp.pi.rsrc into myproj.pi.rsrc,
everything was fine. Any help would be appreciated.

Brian

[e-mail]











Path: ucivax!gateway
From: [e-mail] (Brian Williams)
Subject: RE_Dagnabit_ResFIles
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 11
Date: 11 Jul 93 15:51:10 GMT


"Where there's a whip, there's a way?" No,
"Where there's a whip, there's a dimwit."

I forgot to reset the port to the printer after those frappin error
messages.

I'd like to thank Jim Wintermyre for his responce.

Brian











Path: ucivax!gateway
From: [e-mail] (Jim Lynch)
Subject: Re:  Think C 6.0
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 24
Date: 12 Jul 93 09:46:30 GMT

Hey Symantec! This one drives me up the wall!!

[e-mail] sed:
>>>>>
When I open a project file with it, it does not set the directory that standard
file uses to the directory with the project in it.
<<<<<

Now, I _do_ have uses for getting the think tree, but I want my source files
to always be available easily.

Here's a suggested solution: Have a pair of radio buttons in a custom file
open box. One's title is "Project Tree" and the other is "Compiler Tree".
The project tree radio is the one selected after opening a project, and when
opening a file, the directory displayed is the one containing the prevously
opened project. When navigating, things happen as usual. When somehow
dismissing the dialog, it remembers the directory for next time as usual.

When the user selects the other radio, it saves the directory location into
a space designated and displays the Think tree.

Simpler suggestion (that might be less confusing: Same idea, nothing gets
saved and the radios are standard buttons that work like the 'Desktop'
button.










Path: ucivax!gateway
From: [e-mail] ("Jason A. Bobier")
Subject: missing functions
Message-ID: <[e-mail]>
X-Mailer: ELM [version 2.3 PL11]
Newsgroups: fa.think-c
Lines: 28
Date: 12 Jul 93 14:04:15 GMT

Hi, before I declare a bug in the iostreams for Symantec C++ perhaps someone
can tell me where to find the functions hex() and oct() which according to
Stroustrup are supposed to be declared in the iostream.h.

I was trying to implement a program in his book that went like this, and all
I could get was errors.

#include <stream.h>
#include <string.h>

main()
{
    char chdr = 'a';

    cout << int(chdr) << " = "
        << oct(chdr) << " = "
        << hex(chdr) << "\n";

}

This is actually a shortened version of the program. It is in chapter 2 I
think. In any case, the int() function works but the other two don't.

Thanx,
Jason A. Bobier
[e-mail]

'Living is easy with eyes closed' -The Beatles










Path: ucivax!gateway
From: [e-mail] (Troy Anderson)
Subject: missing functions
Message-ID: <[e-mail]>
Content-Type: TEXT/PLAIN; charset=US-ASCII
Mime-Version: 1.0
Newsgroups: fa.think-c
Lines: 31
Date: 13 Jul 93 06:12:28 GMT

[e-mail] wrote:
>>>>>
perhaps someone can tell me where to find the functions hex() and oct()
which according to Stroustrup are supposed to be declared in the iostream.h.

        cout << int(chdr) << " = "
                << oct(chdr) << " = "
                << hex(chdr) << "\n";
<<<<<
oct and hex are included with the Symantec C++ compiler.  They are defined
to take and return a reference to an iostream, and are declared as follows:

ios& oct(ios&);
ios& hex(ios&);

so you can use them like this:

cout << oct << chdr << " = "
     << hex << chdr << "\n";

as for the int(chdr) part, that creates a temporary int and initializes it
to have the value of chdr.  That is a result of C++'s built in constructor
for the int class.

Section 10.4.2.1 Standard I/O Manipulators in Stroustrup shows an example
of this.

-Troy













Path: ucivax!gateway
From: [e-mail] (Stefan Arentz)
Subject: Summary: Popup Menus and _TrackControl
X-Sender: [e-mail]
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 35
Date: 14 Jul 93 14:08:01 GMT

The Problem:
------------

If you install a popup menu control, using the standard System Seven CDEF
(ID = 1008), then the following code doesn't pop up the menu:

     thePart = TrackControl(theControl, localMouse, nil);

The code does work for Buttons, Radio Buttons and Checkboxes.

The Solution:
-------------

MPW 411 says:

     "If actionProc is POINTER(-1), TrackControl looks in the control record
     for a pointer to the control's default action procedure. If that field
     of the control record contains a procedure pointer, TrackControl uses
     the action procedure it points to; if the field contains POINTER (-1),
     TrackControl calls the control definition function to perform the
     necessary action. (If the field contains NIL, TrackControl does nothing.)"

So, if you pass (ProcPtr)-1 to the actionProc parameter of _TrackControl
everything works fine.

     thePart = TrackControl(theControl, localMouse, (ProcPtr)-1);

Yeah! Nasty!

-- Stefan


----------------------------------------------------------------------------
Stefan Arentz          -  Software Constructor  -          [e-mail]











Path: ucivax!gateway
From: [e-mail] (Stefan Arentz)
Subject: Writing ResEdit Editors
X-Sender: [e-mail] (Unverified)
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 17
Date: 14 Jul 93 21:03:44 GMT

Hello World,

I am writing a ResEdit editor to edit the XDTL resources of my Dialog
Manager Library. I am having some problems with testing (not debugging) the
editor. If I want to test it, I have to build a code resource, quit
ResEdit, copy the resource to the 'ResEdit Preferences' file, start ResEdit
and finally run the editor.

Is there an easier way to do this ?!

BTW: If this is editor is finished I will write a better DITL editor.

-- Stefan

----------------------------------------------------------------------------
Stefan Arentz          -  Software Constructor  -          [e-mail]











Path: ucivax!gateway
From: [e-mail]
Subject: File Control Blocks
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 12
Date: 15 Jul 93 18:45:11 GMT

Hello,

Heres questions for you all:

I have an application which displays the open files on a Macintosh volume.
Does anyone know of a way of finding the process which has opened a file?
Currently I scan the open files using PBGetFCBInfo.  The FCBPBRec returned
gives no indication of the process which opened the file.

Thanks in advance.
-Mark











Path: ucivax!gateway
From: [e-mail] (Big Nose)
Subject: cdev runner
Via: UK.AC.AFRC; 19 JUL 93 12:09:35 BST
Via: uk.ac.afrc.iape; Mon, 19 Jul 1993 12:13:20 +0100
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 8
Date: 19 Jul 93 11:07:35 GMT


Is there a system 7.0 version of the cdev runner project available? The
version supplied with THINK C 5.0 doesn't work, and I can't find an
update in the think-c directory. Anyone know of a source?

Andy Law

(LAWA @ AFRC.AC.UK                       Big Nose in Edinburgh )










Path: ucivax!gateway
From: [e-mail] (pfterry)
Subject: Think C Project Directory Hack
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 12
Date: 20 Jul 93 05:46:33 GMT

A number of people have complained about the behavior of Think C's standard
open and save dialog boxes. Dave Heller wrote a hack and posted it to AOL.
Here's his description.

This is a very simple INIT which hacks the THINK C 6.0 editor (THINK Project
Manager) to default to the project directory like it did for THINK C 5.0.x.  I
found the new default behavior annoying, so I fixed it.  It does not modify the
THINK Project Manager application in any way. Let me know if it doesn't work
for you.


I've placed it in the incoming directory on the ThinkC site.










Path: ucivax!gateway
From: [e-mail]
Subject: Problem with dead code stripping in MPW C++
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 77
Date: 20 Jul 93 08:39:39 GMT
X-Envelope-to: [e-mail]

I know this isn't a Symantec problem but I havn't got my upgrade yet so I
only have MPW C++ and it may be a problem in Symantec C++ too, so here
goes...

I am declaring global instances of derived classes and the linker appears
to be removing them.

The code below compiles and runs as expected under Borland C++ on DOS but under
MPW on Macintosh the linker strips out the classes and the globals never
get created!

The following is the code I'm using. _Any_ ideas on this would be appreciated.

Cheers,

Duncan Lennox

// Start of file "base.h"
class Base
  {
    public:
      static Base* head;
    Base* next;
    Base( void )
      {
        next = head;
        head = this;
      }
    virtual ~Base( void )
      {
        head = next;
      }
    virtual void MyNameIs( void ) = 0;
    static void ListAll( void );
  };

// Start of "base.cp"
#include "base.h"

Base* Base::head = 0;

void Base::ListAll( void )
  {
    Base* b = head;
    while ( b != 0 )
      {
        b->MyNameIs();
        b = b->next;
      }
  }

// Start of file "derived.cp"
class derived : public Base
  {
    public:

      void MyNameIs( void )
        {
          cout << "derived" << endl;
        }
  } TheDerived;

// This is what MacApp's macroDontDeadStrip would expand to (almost)
derived* dummy = new derived;

// Start of file "main.cp"
#include "base.h"

int main( void )
  {
    Base::ListAll();
    return 0;
  }
+----------------------------------------------------------------------------+
Duncan Lennox, Computer Science, U.C.D., Belfield, Dublin 4, Ireland
[e-mail]     "The best way to accelerate Windows is at 9.81ms^-2"











Path: ucivax!gateway
From: [e-mail]
Subject: Re: Problem with dead code stripping in MPW C++
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 22
Date: 20 Jul 93 09:17:23 GMT
X-Envelope-to: [e-mail]

>I am declaring global instances of derived classes and the linker appears
>to be removing them.

Typically having spent the guts of a day at this problem and gotten
nowhere, as soon as I posted a query to the net I found the answer myself!

If I include a "#pragma force_active on" at the start of the source code
that is getting stripped then the linker leaves it alone. This solves my
problem perfectly and the code now works as expected.

A case of RTFM, but its kinda buried and hard to find.

I would still be interested in hearing any other advice or info on this
though, but my panic is over.

Cheers,

Duncan
+----------------------------------------------------------------------------+
Duncan Lennox, Computer Science, U.C.D., Belfield, Dublin 4, Ireland
[e-mail]     "The best way to accelerate Windows is at 9.81ms^-2"











Path: ucivax!gateway
From: [e-mail] ("Sidney W. Kitchel")
Subject: Bug list
Message-ID: <[e-mail]>
In-Reply-To: <[e-mail]> from "[e-mail]" at Jul 20, 93 08:39:39 am
X-Mailer: ELM [version 2.4 PL0]
Content-Type: text
Content-Length: 495
Newsgroups: fa.think-c
Lines: 11
Date: 20 Jul 93 14:05:29 GMT

    I'm a novice at this game having just received my copy of
SC++. Would someone kindly send me mail about where I can get a copy
of the bug list, which was mentioned on one of the comp.sys.mac.* news
groups?
                        Thanks,
                            --Sid
--
Sidney W. Kitchel          [e-mail], [e-mail]
Data Parallel Systems, Inc.  ============||  DPSI  ||===============
4617 E. Morningside Drive                          [phone]
Bloomington, Indiana,  47408  USA        FAX:  [phone]










Path: ucivax!gateway
From: [e-mail] (Mark Nagel)
Subject: ARCHIVE: Marker Control 0.60b
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Reply-To: [e-mail]
Lines: 34
Date: 20 Jul 93 16:31:54 GMT

This is a new version of Marker Control, a control panel that causes
files in THINK C to be marked and their window positions saved.

NEW VERSION v0.6b
  % Fixed a bug where if you had the "Only Mark Files Newer Than:"
    option OFF and moved a window without making any changes to it,
    it set the modified date of the associated file to some time
    in the early 1900s.
  % Made it so that window positions and sizes are saved even if
    you close the windows with "Close All" or by quitting THINK C.
    Before, they were only saved if you actually used "Close" from
    the File menu.
  % Now functions that contain function prototypes as parameters are
    marked.

Executive Summary
-----------------
  % Save window positions in THINK C.
  % Mark functions in source files.
  % Drop RMarker ControlS onto your System Folder and restart.
        Open the RMarker ControlS control panel to configure its operation.
  % Run THINK C and notice the difference.
      - Window positions are saved (open a file, move its window,
        close it, reopen it to try this out)
      - Command-click on the title bar of source file windows to
        see a list of the functions in that file.
  % #pragma mark <yourLabel>    // to mark things other than functions
  % #pragma markOff             // turn marking off until the end of the file or a #pragma markOn
  % #pragma markOn              // turn marking back on

Brought to you by Troy Anderson
[e-mail]

[saved as: /mac/think-c/compiler/marker-control-06b.hqx; 22K]










Path: ucivax!gateway
From: [e-mail]
Subject: 3d library
Message-ID: <[e-mail]>
Content-transfer-encoding: 7BIT
MIME-version: 1.0
Newsgroups: fa.think-c
X-VMS-To: APO"[e-mail]"
Organization: University of Texas Health Science Center at San Antonio
Lines: 1
Date: 20 Jul 93 19:44:55 GMT
X-Envelope-to: [e-mail]

has anyone used the 3d graph library that comes with think-c? i'm looking for source code examples that uses the library.










Path: ucivax!gateway
From: [e-mail]
Subject: default window to Printf
Message-ID: <[e-mail]>
Content-transfer-encoding: 7BIT
MIME-version: 1.0
Newsgroups: fa.think-c
X-VMS-To: APO"[e-mail]"
Organization: University of Texas Health Science Center at San Antonio
Lines: 1
Date: 20 Jul 93 20:39:06 GMT
X-Envelope-to: [e-mail]

i would like to use the PRINTF function to display some variables without havingto create a window first. does think-c provides a default window in which to work in. if so what is the command to display the window, i.e FOPEN(*,"rw")...)?










Path: ucivax!gateway
From: [e-mail] (pfterry)
Subject: RE: Think C Project Directory Hack
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 27
Date: 21 Jul 93 14:02:18 GMT

Lindsay and others,

>> A number of people have complained about the behavior of Think C's standard
>> open and save dialog boxes. Dave Heller wrote a hack and posted it to AOL.
>> Here's his description.
>>
>> This is a very simple INIT which hacks the THINK C 6.0 editor (THINK Project
>> Manager) to default to the project directory like it did for THINK C 5.0.x.

>> I
>> found the new default behavior annoying, so I fixed it.  It does not modify
>> the
>> THINK Project Manager application in any way. Let me know if it doesn't work
>> for you.
>>
>>
>> I've placed it in the incoming directory on the ThinkC site.

>Hmm, I obviously am looking in the wrong place, _which_ site are you referring
>to?

Sorry I was so vagure. You'll find it on ics.uci.edu in the directory
/mac/thinkc/compiler/ and it's name is TPMSF_Hack.sit.

Fred Terry












Path: ucivax!gateway
From: [e-mail] (Troy Anderson)
Subject: Re: default window to Printf
Message-ID: <[e-mail]>
In-Reply-To: <[e-mail]>
Content-Type: TEXT/PLAIN; charset=US-ASCII
Mime-Version: 1.0
Newsgroups: fa.think-c
Lines: 22
Date: 21 Jul 93 16:01:59 GMT


On 20 Jul 1993 [e-mail] wrote:

> i would like to use the PRINTF function to display some variables
without havingto create a window first. does think-c provides a default
window in which to work in. if so what is the command to display the
window, i.e FOPEN(*,"rw")...)?
<

You don't have to call anythingh before printf.  As long as you have ANSI
in your project, just call printf.  The first time you call anything that
sends something to stdout or stderr or requests something from stdin, the
window is opened and you're in business.  A word of caution, it's not a
good idea to use standard quickdraw calls while you are using the ANSI
window.  That means, you either do all the work yourself - NewWindow,
DrawString, etc, or you let the ANSI library do all the work - printf,
etc.  You can, however, use sprintf any time you like.

Hope this helps,
Troy












Path: ucivax!gateway
From: [e-mail]
Subject: Serial Printing
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 11
Date: 21 Jul 93 21:42:58 GMT


Does anyone know of a printer driver for System 7 that prints in plain old-
fashioned ascii, without any control characters, like an old line-printer
on the VAX or PC?

Also, anyone with tips about converting a non-mac serial printer to mac use
would be greatly appreciated.

[e-mail]
Fordham University
Bronx,  New York










Path: ucivax!gateway
From: [e-mail] (Danny Thomas)
Subject: Re: Serial Printing
X-Sender: [e-mail]
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 46
Date: 21 Jul 93 22:41:11 GMT

>Does anyone know of a printer driver for System 7 that prints in plain old-
>fashioned ascii, without any control characters, like an old line-printer
>on the VAX or PC?
>
>Also, anyone with tips about converting a non-mac serial printer to mac use
>would be greatly appreciated.
You could look at Chuck's Printer Drivers on Sumex-aim which support
Epson-type printers in both 'ASCII' and graphics modes. I have played
around a little with it but have noticed the following problems which I
haven't yet reported to Chuck. System 7.1 on an LC connected to Epson FX80:
(1) when printing from Word 4.0d (haven't tried other versions), you seem
to get a crash when operating in the graphics modes. The text mode seems
fine, except you have to set narrower margins so lines don't wrap
(2) it doesn't seem to work from within VersaTerm 4.xx (not sure of our
version right now). This is our real need cause we run a terminal session
to an accounting system and want it print out order forms (impact printer
for carbon copies). We have set up VersaTerm to use the Chosen printer, and
Finder prints OK to the Epson. No problems are reported but nothing comes
out of the printer, whether or not Chuck's print spooler is active.


cheers,
Danny Thomas.

PS I used Archie to locate Daisy 1.1 but I can't remember where from
from info-mac (sumex-aim.stanford.edu) but mirrored around the world:
     65394 17 Jun 93 ./prn/chucks-printer-driver-110.hqx



================= EXCERPT FROM DOC FILE ==============================
What?
"Chuck's Printer Driver" is a printer driver -- chooser device for using
Epson dot-matrix printers (and compatibles) with your Macintosh computer.
It has been tested on a Macintosh SE and a Macintosh IIci (systems 6.0.5
through 7.1, English and Kanji versions), with a Panasonic KX-P1080i
printer.  It prints in Text (draft or NLQ) mode and Graphics (Low, Medium,
and High) modes.  It is Freeware, however I do request a small donation
($20 U.S.) to help pay for my costs.  I would like to hear what you think
of it, and what I should add, even if you don't send me money.  Note: There
are still bugs in the text printing option,  but the graphics modes should
work fine. (I really don't want to support text printing that much, there
is a better public-domain printer driver already that does a good job of
that called "Daisy".)












Path: ucivax!gateway
From: [e-mail] (Steve Morris)
Subject: Re: Serial Printing
Message-ID: <[e-mail]>
In-Reply-To: <[e-mail]> ([e-mail])
Newsgroups: fa.think-c
Lines: 14
Date: 21 Jul 93 23:09:10 GMT


PowerPrint connects just about any printer to a Mac serial port. It actually
connects to a parallel port on the printer. The conversion is done in the
cable connector. It's cute and I have been very satisfied with mine connected
to a Lexmark 4029 with PCL 5 cartridge (our code in the cartridge of course.)
It basically connects anything which is compatable with Epson, Proprinter, HP
and something else that I don't recall.

Steve Morris
<[e-mail]>
Phoenix Technologies LTD
38 Sidney Street
Cambridge, MA 02135
[phone]










Path: ucivax!gateway
From: [e-mail] (pfterry)
Subject: FW: Think C Project Directory Hack
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 12
Date: 22 Jul 93 04:34:12 GMT


> Sorry I was so vagure. You'll find it on ics.uci.edu in the directory
> /mac/thinkc/compiler/ and it's name is TPMSF_Hack.sit.

For those of you with automatic file-retrieval software, that's:

    ics.uci.edu:/mac/think-c/compiler/TPMSF_Hack.hqx

--
 Pete Gontier // EC Technology // [e-mail]

forwarded










Path: ucivax!gateway
From: [e-mail] (Mark Nagel)
Subject: ARCHIVE: File Dropper 1.1b2
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Reply-To: [e-mail]
Lines: 52
Date: 22 Jul 93 05:54:23 GMT

File Dropper 1.1b2
Written by Troy Anderson

Copyright (c) 1992-1993, Troy Anderson; All Rights Reserved

What Is This Thing?

File Dropper is a THINK C 6.0 library that implements an application that you
can drop files or folders onto to do batch operations on.  It handles the
getting of the AppleEvents if running under System 7, the main event loop, and
the menus.  You need only write the code that acts on the individual files.

New with version 1.1b2:
o  Built "File Dropper " in THINK C 6.0
o  The cursor is NOT reset to an arrow after each call to SetStatusPercentage.
o  If you return FALSE from the eStartup, eSFInitialize, or eAEInitialize
   messages, File Dropper  will send you an eQuitting message and quit the
   application.
o  If you return FALSE from the eValidate message, File Dropper will skip any
   remaining files in the current batch.
o  If you return FALSE from an eDispose message, File Dropper will send you an
   eQuitting message and quit the application.  Returning TRUE from an eDispose
   message will cause File Dropper to continue running.  This behavior does NOT
   depend upon which of eSFInitialize or eAEInitialize was sent to you.
o  If you return TRUE from an eQuitting message, the File Dropper quits.  On the
   other hand, if  you return FALSE from an eQuitting message, File Dropper
   will quit.
o  If you want the source code to File Dropper, send me money (see the end of
   this document).
o  Added an eUserCancelled message for when the user clicks Cancel in the status
   box.
o  Added an eIdle message.
o  Added the ChangeStatusMessage function that lets you change the message in the
   status dialog on the fly, during your processing of the file.
o  Changed some things internally, so you can NOT call
   InstallCustomGetFSSpecFunc, InstallCustomFileFilterFunc, or
   InstallCustomDialogHookFunc from an eSFInitialize or  eAEInitialize message
   any more, just from eStartup.
o  Fixed a bug in the ErrorAlert function so that text relating to the error as
   well as the word "Error." show up in the dialog.
o  Made it simpler to support AboutBox functionality.  You just put text in STR#
   resource 129, and it will be displayed in the warp stars.  To enable this
   feature, return TRUE from the  eDoAboutBox message.


New with version 1.1b1:
o  You specify if you want the status dialog displayed while you are
working on the file. Progress bar like Finder 7's that you update by
specifying how far you are along with a percentage (0 - 100).  You
specify if you want this, too.

[saved as: /mac/think-c/code/file-dropper-11b1.hqx; 70K]










Path: ucivax!gateway
From: [e-mail] (Mark Nagel)
Subject: ARCHIVE: TPM SF Hack
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Reply-To: [e-mail]
Lines: 15
Date: 22 Jul 93 05:55:35 GMT

From: [e-mail] (pfterry)
Subject: Think C Project Directory Hack
Date: 20 Jul 93 05:46:33 GMT

A number of people have complained about the behavior of Think C's standard
open and save dialog boxes. Dave Heller wrote a hack and posted it to AOL.
Here's his description.

This is a very simple INIT which hacks the THINK C 6.0 editor (THINK Project
Manager) to default to the project directory like it did for THINK C 5.0.x.  I
found the new default behavior annoying, so I fixed it.  It does not modify the
THINK Project Manager application in any way. Let me know if it doesn't work
for you.

[saved as: /mac/think-c/compiler/TPMSF_Hack.hqx; 17K]










Path: ucivax!gateway
From: [e-mail] ("Edmund J. Sutcliffe")
Subject: Re: Serial Printing
Message-ID: <[e-mail]>
In-Reply-To: <[e-mail]>
Content-Type: TEXT/PLAIN; charset=US-ASCII
Mime-Version: 1.0
Newsgroups: fa.think-c
Lines: 28
Date: 22 Jul 93 07:19:50 GMT


There is a printer app called daisy which does just that.
    Edmund
===============================================================================
Edmund J. Sutcliffe                                   Phone [phone]
Computing Service,                                    Fax   [phone]
University of York,
Heslington,                                  Internet Mail: [e-mail]
York, YO1 5DD U.K.



On 21 Jul 1993 [e-mail] wrote:

>
> Does anyone know of a printer driver for System 7 that prints in plain old-
> fashioned ascii, without any control characters, like an old line-printer
> on the VAX or PC?
>
> Also, anyone with tips about converting a non-mac serial printer to mac use
> would be greatly appreciated.
>
> [e-mail]
> Fordham University
> Bronx,  New York













Path: ucivax!gateway
From: [e-mail]
Subject: The novel goes on (Was Re: Help reading data from file)
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 190
X-Mts: smtp
Date: 22 Jul 93 17:44:14 GMT

    First, thanks for the sample code Johan!  That helped me to
begin to understand once more how C works.  I wrote my code (well,
*most* of the routine to read/convert/write files), but when I try
to run it, its spits out garbage.  Here is the code:


------------------------------ File convert2SI.c --------------------------
#include "convertSI.h"
#include <stdio.h>
#include <ctype.h>

/**************************************************************************
 *
 *    Function that reads in the contents of the file inFile, converts it to
 *    SI, and writes the result to outputFile.
 *
 *    Parts based (stolen?) on sample code provided by Johan Strandberg.
 *
 *    Last Modified in 21/07/1993
 *
 **************************************************************************/
int convert2SI (inputFile, outputFile)

char *inputFile;
char *outputFile;

{
    FILE    *in;        /* File pointer to the input file */
    FILE    *out;        /* File pointer to the output file */
    int        err;        /* To store temporary error values */
    int        lettre;        /* Make sure it's an int! */
    int        status;
    char    OutString[100];    /* Place to put the strings */
    int        StringCnt = 0;        /* keeps track to the
position in OutString */

     /*
     * Open InputFile
     */
    errno = 0; /* All OK */
    in = fopen(inputFile, "r");
    if (errno) {
        return errno;
    }

    /*
     * Open OutputFile
     */
    errno = 0;
    out = fopen(outputFile, "w");
    if (errno) {
        err = errno; /* save errno since fclose will scramble it */
        fclose(in);
        return err;
    }


     /* Set flag to (char, lowercase) */
     status = CharLower;

    /*
     * Clear string OutString.
     */
    /* While there is something to be read from InputFile */
    while ((lettre = fgetc(in)) != EOF) {

         /* Is lettre character (a-z A-Z)? */
        if (isalpha(lettre)) {
             /* Is flag not set to Char*? */
             if (status <10) {
                 /* Call WriteString to save contents of
OutString to OutputFile. */
                 StringCnt = WriteString(OutString, out);
                 status = CharLower; /* Reset flag to CharLower */
             } /* End if status < 10 */

              /* Is lettre uppercase? */
              if (isupper(lettre)) {

                 if (status == CharLower) { /* Is flag
set to char/lowercase? */
                    status = CharUpper; /* Set flag
to (char/uppercase) */

                    /* Write OpenUpper character to OutString */
                    OutString[StringCnt] = OpenUpper;
                    StringCnt++;

                 } /* End If status == CharLower */

                 /* Write lettre (lowercase) to OutString */
                 OutString[StringCnt] = tolower(lettre);
                 StringCnt++;

             } else if (islower(lettre)) { /* (lowercase) */
                 if (status == CharUpper) { /* Is flag
set to char/uppercase? */
                     status = CharLower; /* Set flag
to (,lowercase) */

                     /* Write CloseUpper character
to OutString */
                     OutString[StringCnt] = CloseUpper;
                    StringCnt++;

                 } /* End If status == CharUpper */

                 /* Write lettre  to OutString */
                 OutString[StringCnt] = lettre;
                 StringCnt++;

             } /* End If case */

        } else if (isdigit(lettre)) { /* is lettre number (0-9)? */

             if (status > 1) { /* Is flag NOT set to (number)? */
                 /* Call WriteString to save contents of
OutString to OutputFile. */
                 StringCnt = WriteString(OutString, out);
                 status = CharNumber; /* Set flag to (number) */

             } /* End If status > 1 */

             /* Write lettre to OutString */
             OutString[StringCnt] = lettre;
             StringCnt++;

        /* is lettre space, ",", etc? */
        } else if (isspace(lettre) || ispunct(lettre)) {

            /* Call WriteString to save contents of
OutString to OutputFile */
            StringCnt = WriteString(OutString, out);

            /* Write lettre to OutputFile */
            fputc (lettre, out);

        } /* End If */

    } /* End While */

    /* Close opened files */
    fclose(in);
    fclose(out);

    return 0;

}

int WriteString (outstring, outfile)
char    *outstring;
FILE    *outfile;
{
    int    last_one = strlen(outstring);    /* how many characters are *

     * in the string?          */
    int    i;
    /* Counter for "for" loop */

    if (last_one >0) { /* work only for non-empty strings */
        for (i=last_one; i>=0;i--) {
            fputc (outstring[i], outfile);
        }
    }

    return 0; /* reset counter */


}

------------------------------ File convert2SI.c --------------------------

The header file convertSI.h has only the values I am using for the
variables CharUpper, CharLower, CharNumber, OpenUpper, and OpenLower.
Below is the test file I created


------------------------------ File test.in --------------------------
This is a TEst.?>  UUUU.uuu-123
------------------------------ File test.in --------------------------

And here is what I got back:

------------------------------ File test.out --------------------------
XDNIsih}t{ XDNIsih}si XDNIsih}sa
XDNIts}et{.XDNIts}et{?XDNIts}et{>XDNIts}et{ XDNIts}et{
XDNItuuuu{.XDNItuuuu}-XDNItuuuu}
------------------------------ File test.out --------------------------

What have I done wrong?










Path: ucivax!gateway
From: aadt![e-mail] (Atul Barve)
Subject: Managing versions of c files
X-Sender: atulb@192.9.225.3
Message-ID: <9307222141.AA00428@tears>
Newsgroups: fa.think-c
Lines: 30
Date: 23 Jul 93 00:00:23 GMT

hi netters,

I have written a simple version control system for think c. it does not
have any of the sofiticated features of source server or Projector.

This is what it does...

It allows multiple people working on a think c project, to coordinate there
development efforts. so if userA is working on file1 and userB wants to
work on the same file it will not allow you to edit it (Provided Projector
Aware flag is set). It does it thru the use of apple events and program
linking. When userA is done editing the file he checks it in. If it has
changed it copies the file to a specified directory. One advantage of this
system is that it allows you to maintain source code locally. and still
keep track of it...

Would such a tool be usefull ???

thanks,





Atul (!Ajay)
([e-mail])

"Don't take life too seriously... it's not like anyone gets out of it alive."
                        -- [e-mail]











Path: ucivax!gateway
From: [e-mail] (Danny Thomas)
Subject: posting messages with enclosures
X-Sender: [e-mail]
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 28
Date: 23 Jul 93 01:04:16 GMT

>Executable binary (application) of Arashi 1.1, a vector graphics game
>that requires 256 colors or shades of gray.
>
>Juri Munkki
>[e-mail]
>
>
>
>Attachment converted: !Download:ARASHI 1.1.sea (Type: 'APPL' Creator: 'DSEA')

From some of the newsgroups I subscribe to I know that numerous people have
well founded objections to people posting messages with (large) enclosures.
Some concerns include bandwidth (particularly people _paying_ for their
modem/SLIP links), possibility of virii, etc. My suggestion is to first
post a message asking whether (a) there are enough people interested in the
enclosure to make it worthwhile (b) not enough people that object, and then
make a decision whether to mail out the enclosure. Personally I think
placing these on an FTP site is a better option leaving it to people who
are interested to fetch.

just my 0.02c worth
Danny Thomas.


PS I'm writing this message mainly because of the size of the enclosure
(263K), and because it doesn't seem related to the purpose of this mailing
list (eg it doesn't seem to include source code).











Path: ucivax!gateway
From: [e-mail] (Mark Nagel)
Subject: ARCHIVE: Arashi 1.1 source
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Reply-To: [e-mail]
Lines: 12
Date: 23 Jul 93 01:44:58 GMT

You probably all already know this :), but the source for Arashi 1.1
is available.  Even though it was mistakenly sent to the entire list
(Juri has already apologized, so don't roast him _too_ badly), I've
stashed a copy in the archives.

I figure the pain of 450+ duplicate copies is probably dulled by the
fact that Juri has made available the entire source for this very cool
game.  Thanks, Juri!

[saved as: /mac/think-c/appl/arashi-11.hqx; 1408K]

Mark










Path: ucivax!gateway
From: [e-mail] ("Jason A. Bobier")
Subject: missing functions (fwd)
Message-ID: <[e-mail]>
X-Mailer: ELM [version 2.3 PL11]
Newsgroups: fa.think-c
Lines: 40
Date: 23 Jul 93 05:16:40 GMT

> perhaps someone can tell me where to find the functions hex() and oct()
> which according to Stroustrup are supposed to be declared in the iostream.h.
>
>         cout << int(chdr) << " = "
>                 << oct(chdr) << " = "
>                 << hex(chdr) << "\n";
> <<<<<
> oct and hex are included with the Symantec C++ compiler.  They are defined
> to take and return a reference to an iostream, and are declared as follows:
>
> ios& oct(ios&);
> ios& hex(ios&);
>
> so you can use them like this:
>
> cout << oct << chdr << " = "
>      << hex << chdr << "\n";
>
> as for the int(chdr) part, that creates a temporary int and initializes it
> to have the value of chdr.  That is a result of C++'s built in constructor
> for the int class.
>
> Section 10.4.2.1 Standard I/O Manipulators in Stroustrup shows an example
> of this.
>
> -Troy

Thanx Troy, but what I'm looking for are the functions mentioned on page 54
of The C++ Programming Language. The sentence describing them reads:
 "The function oct() produces an octal representation of its integer
argument, and hex() produces a hexadecimal representation of its integer
argument; oct() and hex() are declared in <iostream.h>."

There are exercises on page 74 that require these functions. Anyone else
seen these? Phil?

Jason A. Bobier
[e-mail]

'Living is easy with eyes closed' -The Beatles










Path: ucivax!gateway
From: [e-mail] (Jim Lynch)
Subject: Re:  missing functions (fwd)
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 14
Date: 23 Jul 93 09:24:14 GMT

>>>>>
Thanx Troy, but what I'm looking for are the functions mentioned on page 54
of The C++ Programming Language. The sentence describing them reads:
 "The function oct() produces an octal representation of its integer
argument, and hex() produces a hexadecimal representation of its integer
argument; oct() and hex() are declared in <iostream.h>."
<<<<<

I tried to use the hex(int) and oct(int) manipulators but they weren't in
gnu c++. You could write your own... It'll take you awhile to find out how
to write manipulators, however I think there's some treatment of the subject
in Think Reference under iostreams. A good explanation probably also is in
C+C++ by Allen I Holub, a very readible book that goes into considerable
technical detail about C++. A little more than a double sawbuck...










Path: ucivax!gateway
From: [e-mail] (Per Mildner)
Subject: Re: Managing versions of c files
X-Sender: [e-mail]
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
X-Charset: ASCII
Lines: 16
Date: 23 Jul 93 12:12:38 GMT
X-Char-Esc: 29

At 00.00 93-07-23 +0000, Atul Barve wrote:
>
>Would such a tool be usefull ???
Possibly, is it available?
One thing that I would find very useable would be something that does not
require a shared file-system nor network at all times. Eg, we are a few
people working on the same software at home. It would be nice to have to
explicitly check out stuff for modificatin, and to be able to merge or at
least find all modified sources since the last version. Typically the
softare is transported on a powerbook (although not developed on one) so
the merge could require shared AppleTalk.
Per Mildner                      [e-mail]
Computing Science Dept.
Uppsala University               tel: [phone]
Box 311, S-751 05 Uppsala,Sweden fax: [phone]











Path: ucivax!gateway
From: [e-mail] (Sigurdur Darri Skulason)
Subject: bug in comm toolbox interface?
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
X-Charset: ASCII
Lines: 42
Date: 23 Jul 93 18:00:10 GMT
X-Char-Esc: 29

I have an communications program which compiled quite nicely under think c 5.
Recently, after making some modifications I compiled it under thin c 6 and
started to get a number of problems. When the program was run from the think
c environment everythink worked out ok. When compiled as a application however
It always crashed at seemingly random locations during the startup process.

Rebooting with no inits, rebuilding the project, reinstalling the compiler
etc changed nothing and very close expection of the code still fails to show
any bugs anywhere. After elimininating every possible cause for crash and
stripping away large chunks of code, left me with nothing but the basic
toolbox initialization and the code:
if (InitCRM() != cmNoErr)
        FatalError(eNoCRM);
and the system hung inside the InitCRM. That only happened though when
the program was compiled as an application,not when run inside think C.

I then tried reinstalling the comtoolbox library and creating a small project
consisting of commtoolbox, mactraps and a test.c containing:
main() { SysBeep(); }
running that project beeps nicely, but when I tried to compile a application
from it, Yo and behold, during linking I get an "Out of memory error".
Even though think C has 5 mb of memory, no inits installed and running on
an fx.
  Removing the CommToolbox I can create an application without problems.

So I seem to be led to the conclusion that there is something weird about
the comtoolbox library. I have a fresh copy of think c, the library,
no inits/cdevs, sys 7.1 on a mac fx.

Therefore.. the point in this is.. has anyone tried to compile a application
with think c 6.0 using the comtoolbox undir similar circumstances? If so and
it has succeded then does anyone have an idea what might be wrong?
Might some resource in the system be causing this?

Thanks from a confused Sigurdur.

--
//
// Sigurdur Darri Skulason           [email] [e-mail]
// President,                    [applelink] ICE0008
// Hugsun computer services            [tel] [phone]
// Iceland                             [fax] [phone]










Path: ucivax!gateway
From: [e-mail] (Dana S Emery)
Subject: Re: Managing versions of c files
Message-ID: <[e-mail]>
In-Reply-To: Your message <[e-mail]> of 23 Jul 93
 12:12:38 GMT
Content-Type: TEXT/plain; charset=US-ASCII
Newsgroups: fa.think-c
Lines: 17
Date: 23 Jul 93 22:27:10 GMT

>  Would such a tool be usefull ???

maybe, but what I would find even more useful, would be a way to write lock
source files without haveing to check them in/out of a source manager.

I like to fiddle with comments, I also like to browse the TCL source code.
Mixing those 2 is a bad idea.

I have used MPW projector to protect "library" source, both sysmantec and other,
but the megabyte sized archive files which are created when one checks in the
TCL source are a pain to have to keep around.

The solution is simple enough, just add/remove a dummy checkout flag to whatever
files need protection.
--
dana s emery <[e-mail]>











Path: ucivax!gateway
From: [e-mail] (Francois Menneteau)
Subject: Color popup menu problem...
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Organization: IMAG Institute, University of Grenoble, France
Lines: 16
Date: 24 Jul 93 11:49:05 GMT

Thinkc 5.0.4/TCL 1.1.2 any machine/System 7.1

I have a color popup menu in a menu. When I pull down it the first time, its
color are ok. However, when I pull down it the next time, all the colors are
lost. Have you experiment this? Have you solve this problem?

PS: I have solve it by using a 'mctb' with an ID of 0, but this does not allow
real customization...

Thanks in advance.

--
========================== ()__|||||__() "... I had their lives in my hands
[e-mail] () /O   O\ () their fate their fortune in my visions
LGI-IMAG. Tel: [phone] () - .|. - () No one believed in my true prophecy
========================== ()  \=^=/  () And now it's too late."  (Iron Maiden)










Path: ucivax!gateway
From: [e-mail] (Sigurdur Darri Skulason)
Subject: reason for ctb bug
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
X-Charset: ASCII
Lines: 21
Date: 25 Jul 93 17:18:04 GMT
X-Char-Esc: 29

Thanks to those who made suggestions about the reasons for the comm toolbox
crashing.
To sum it up: I got a few replies suggesting that smart linking should be
turned off because it's plainly bugged. I didn't know that so thanks for the
info.

Also someone who had got the same problem with out of memory errors using the
comm toolbox suggested simply to open the comm toolbox library with think c
 and rebuild a new library from it, since that had solved his problem.
That proved also to be the solution and I got rid of TC out of memory alerts and
random crashes in the compiled code.
 So if anyone has strange problems with the communication toolbox which cannot
be blamed upon the phase of the moon, then try that solution.

Thanks,
  Sigurdur.
--
// Sigurdur Darri Skulason           [email] [e-mail]
// President,                    [applelink] ICE0008
// Hugsun inc.                         [tel] [phone]
// Iceland & The Netherlands           [fax] [phone]










Path: ucivax!gateway
From: [e-mail] ("Paul S. Michelsen")
Subject: Objects as resources?
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 38
Date: 26 Jul 93 17:28:11 GMT

I have a large number of objects that I want to store as resources.  I
want the data in each resource to become the values in the object, but
without copying everything if possible.  This will allow me to make
the objects purgible, but easily recoverable with GetResource() (I
think) and what ever conversion process I did the first time.  The
data in the objects are constant.

The objects are instances of several subclasses of a common
super-class.  Each resource needs to identify the subclass to which
the object should be a member.  Part of the conversion process from
resource to object must involve picking the correct subclass.

In Think C 4, I could use the bless() function.  Each object contained
a hidden integer, which is used to hold the class type of that
object.  I used that space in the resource to hold a subclass type ID,
which is a constant defined in an enum{}.  Upon reading the resource I
used this ID in a switch statement, each case of which bless()ed the
resource to be a member of the correct subclass.

Then I got the Think C 5 upgrade for Christmas.  (And it's already
obsolete ;-).  I've started working on that code again, but Think C 5
gives a syntax error on the bless() statement.  I can't find the
bless() function in any of the indexes of the manuels, or listed with
the other oops functions.  Is it gone?  Is there something I can
replace it with?  Or is there a better way to do what I want?

The number of these objects may grow large enough that I might need to
manage them myself, without the direct help of the resource manager.
If this happens then my questions above will be void.

Thanks.

=|<>|=  Paul S. Michelsen                  Eventually, computers will be
=|  |=  [e-mail]              infinitely fast, impossibly small,
=|  |=  ext x23224, rm M2-160                       and free.
=|__|=  Amdahl, Sunnyvale, California
Public service message: The 21st century begins on 1/1/2001, not 2000.











Path: ucivax!gateway
From: [e-mail] (Rolf Staflin)
Subject: Address error when linking with IOSTREAMS
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Reply-To: [e-mail]
Lines: 38
Date: 27 Jul 93 01:39:47 GMT


Bug report:
-----------
If I open the library project IOSTREAM and then choose "Check link"
from the Project menu, I get a system error similar to
"Address Error at 00145ED4 'CODE 0014 0294'+45E0".

My system
---------
Mac Plus, Brainstorm accelerator, 4 Mb RAM, 117 Mb harddisk, StyleWriter.
Multifinder, Finder S1-6.1.7, System S1-6.0.7, MacsBug 6.2.2. No INITs.

Comment
-------
This means that I can't use the library in any projects, which in turn
means that I can't do anything useful without resorting to C. My old C
source (about 80 Kb source, using most of the managers) works fine, though.

  I tried reinstalling, checking for viruses, running on a friend's
Mac SE (same result), cursing and more. After reading of Mr.
Skulason's problem, I tried recompiling the ANSI++ and IOSTREAMS
libraries. Compiling works fine but the linker still goes down in
flames. I even tried creating a fresh project, adding all files that
are in the original IOSTREAMS, copying the options and recompiling.
Same result - when I choose "Check link", I get an address error in
stead of the link errors that I should get.

  I'm surprised that I haven't seen this bug mentioned here before.
Dozens of people should have complained, unless I am doing something
wrong (which I hope, but it seems unlikely) or I've hit on a unique
combination of hard- and software for which the linker doesn't work on
C++ code. Any help would be appreciated.

/Rolf

/* -------------------- */
Rolf Staflin         Roslagsgatan 2 - 4    S-753 27 Uppsala  SWEDEN
Phone: [phone]   Fax: [phone]     [e-mail]










Path: ucivax!gateway
From: [e-mail] (Mike Greenawalt)
Subject: No library manual?
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 28
Date: 27 Jul 93 15:58:21 GMT

I just got my Think C/SC++ upgrade late last week.  I popped open the box and
started poking through the manuals (where on earth did they find that ghastly
yellow color?) looking for goodies.

Suddenly I realized there was something missing!  Where is the library
reference manual?  It is not mentioned in the list of "What is in the package",
so I suppose it was not lost.

Has Symantec decided that the library reference is not needed any longer?  I am
surprised.  For me, the library reference was/is one of the more frequently-
used pieces of the former packages.

It's all well and good that standard ANSI library components are probably
documented in other places, but what about the non-ANSI stuff such as the
console library?  What about the unix-like inclusions?  These are not so
"standard".

Can I assume that my library reference for version 5 still holds true?  I have
Think C manuals dating back to version 1.  What does the first-time buyer do
today?

+---------------------------+--------------------------------------------+
| Mike Greenawalt           | [e-mail]                           |
| Metaphor Computer Systems | Metaphors be with you.                     |
| 1965 Charleston Road      |                                            |
| Mountain View, CA 94043   | These opinions are mine, not my employer's |
+---------------------------+--------------------------------------------+











Path: ucivax!gateway
From: [e-mail] (Rich Siegel)
Subject: Re:  No library manual?
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 4
Date: 27 Jul 93 17:15:56 GMT

The documentation for the standard libraries is now in a THINK Reference
database. Lowers their COGS, and makes it easier to look stuff up...

R.










Path: ucivax!gateway
From: [e-mail] (Michael Crawford)
Subject: Re:  No library manual?
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 21
Date: 27 Jul 93 18:17:08 GMT

Read the source code, of course!  Lots of people don't realize it's there,
and once they see ThinkC's ansi implementation, it scares them so much they
either write their own, or stop using it.

For a good time, look at the code for atoi - it calls sprintf, instead of
StringToNum.  Also, the startup code for stdio patches ExitToShell - so you
will crash if you use it in a code resource, so be sure to comment out
the patch if you are really going to use it.

It is helpful to go through the code and create markers for all the functions.
Also, instead of using libraries, just include the relevant sources in your
program - it makes it much easier to debug, because you can break in the
library source within the source debugger.

Enjoy!

Michael Crawford   | UC Santa Cruz, visiting CERN
Division PPE       |
CERN               | [e-mail]
1211 Geneva 23     |
Switzerland        | "Tilting at windmills for a better tommorrow."










Path: ucivax!gateway
From: [e-mail] (Jerry Wilcox)
Subject: Re: No library manual?
X-Sender: [e-mail]
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 20
Date: 27 Jul 93 18:26:38 GMT

Mike Greenawalt writes:
>Suddenly I realized there was something missing!  Where is the library
>reference manual?  It is not mentioned in the list of "What is in the package",
>so I suppose it was not lost.
>

No, it was a deliberate omission. Think C 6.0 (and C++) ships with a copy
of THINK Reference 2.0.1 and a database describing the standard C
Libraries. All you have to do is put an alias to Think Reference into TPM's
Tools folder and name it "THINK Reference". When you option-double-click on
the name of any standard C function, TPM will automatically bring up the
reference. If you already have (or you purchase) THINK Reference, you have
the other databases, so you can, for example, do a similar
option-double-click on a ToolBox call to bring up the reference for that.

Jerry
-----
Jerry Wilcox - [e-mail]  or [e-mail]
All opinions are mine alone - the University can speak for itself.











Path: ucivax!gateway
From: [e-mail] (Wally Wedel)
Subject: Re - No library manual?
X-Sender: wwedel@saguaro
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 32
Date: 27 Jul 93 18:32:10 GMT

Mike Greenawalt <[e-mail]> writes:

>Suddenly I realized there was something missing!  Where is the library
>reference manual?  It is not mentioned in the list of "What is in the package",
>so I suppose it was not lost.

In the user manual there is a vague reference to a coupon in the back which
can be used to order it. The coupon is missing. I called Symantec about
this and they assured me that no harcopy manual for the standard library is
available.

>Has Symantec decided that the library reference is not needed any longer?  I am
>surprised.  For me, the library reference was/is one of the more frequently-
>used pieces of the former packages.

They replaced it with the Think Reference online database. The database
works pretty well, but I still miss the paper copy and often use the
Version 5 manual.

>Can I assume that my library reference for version 5 still holds true?  I have
>Think C manuals dating back to version 1.  What does the first-time buyer do
>today?

The version 5 manual seems to be OK, but I double check stuff with the
online version to be sure.

Wally Wedel
U S WEST Communications
4001 Discovery Drive, Suite 370
Boulder, CO 80303-7813
Internet: [e-mail]  AppleLink: D5100   Voice: 303-541-6052











Path: ucivax!gateway
From: [e-mail] (Mike Greenawalt)
Subject: Summary:  No library manual?
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 13
Date: 27 Jul 93 18:49:26 GMT

Thanks for all the many replies.

Many folks pointed out that the library references are now on-line.  I had not
yet attempted to install the package yet, so have not run across the files.
The information is slightly buried in the manual.

I called Symantec to inquire about the hard copy of the library manual that is
mentioned in the User's Guide.  The library manuals are not ready yet, but
they will take your name.

For myself, I like to have the book in my hands.

-- Mike










Path: ucivax!gateway
From: [e-mail]
Subject: No library manual?
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 12
Date: 27 Jul 93 19:31:04 GMT

Mike Greenawalt asked where the Standard Library Manual was for Symantec C++
...

It's on disk and gets installed to be accessed by the THINK REFERENCE
application.

You can argue that the manual was nice, but having it handy online is also
nice.

Linda
[e-mail]











Path: ucivax!gateway
From: [e-mail] (Kathleen Fischer)
Subject: Re:  No library manual?
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
X-VMS-To: addvax::in%"[e-mail]"
Lines: 6
Date: 27 Jul 93 19:40:33 GMT
X-Envelope-to: [e-mail]

> The documentation for the standard libraries is now in a THINK Reference
> database. Lowers their COGS, and makes it easier to look stuff up...

Don't you have to pay for this database?

Kathleen










Path: ucivax!gateway
From: [e-mail] (James Ashton)
Subject: Re:  No library manual?
X-Face: rUK3oMwdIS>EJ!*~|~H$SL4S%-}pU+OhHV"`Q}$Dc<~`Tn"N{RjEJV&9;mM}`@P>F;{+]^2
     gO,tx/g0M0[tNFyGMNEjNEyoz06a>{r6Ez3Km}'RN`CpF"$[PmCzGD_}|N\*f2a'EOdg-4h
     7VIL,<0@:t;C/j`^X@_=ckuloh`sGE[eU&W=A#ge3~+"rq9F,r6=9y!V;m};\t],cbJ/|kM
     5ruHFZfE(]Z}jflI%K9O/A*d{O]'3Jz
Message-ID: <[e-mail]>
X-Mailer: ELM [version 2.3 PL11]
Newsgroups: fa.think-c
Lines: 19
Date: 28 Jul 93 01:56:56 GMT

>From: Kathleen Fischer <[e-mail]>
>Subject: Re:  No library manual?
>Date: 27 Jul 93 19:40:33 GMT
>
>> The documentation for the standard libraries is now in a THINK Reference
>> database. Lowers their COGS, and makes it easier to look stuff up...
>
>Don't you have to pay for this database?

I second this question.  I have the version 5 to version 6 upgrade
offer but reading it very carefully gives me the impression that the
new version of Think Reference is not included.  The local distributors
want AUD 60.00 (about USD 41.00) for it.
--
James Ashton                          System Administrator
VK1XJA                         Department of Systems Engineering
Voice [phone]      Research School of Physical Sciences and Engineering
FAX   [phone]                Australian National University
Email [e-mail]               Canberra ACT 0200 Australia










Path: ucivax!gateway
From: [e-mail] (Trey Campbell)
Subject: Re: No library manual?
Message-ID: <[e-mail]>
In-Reply-To: <[e-mail]>
Content-Type: TEXT/PLAIN; charset=US-ASCII
Mime-Version: 1.0
Newsgroups: fa.think-c
Lines: 25
Date: 28 Jul 93 04:20:08 GMT



On 28 Jul 1993, James Ashton wrote:

> >From: Kathleen Fischer <[e-mail]>
> >Subject: Re:  No library manual?
> >Date: 27 Jul 93 19:40:33 GMT
> >
> >> The documentation for the standard libraries is now in a THINK Reference
> >> database. Lowers their COGS, and makes it easier to look stuff up...
> >
> >Don't you have to pay for this database?
>
> I second this question.  I have the version 5 to version 6 upgrade
> offer but reading it very carefully gives me the impression that the
> new version of Think Reference is not included.  The local distributors
> want AUD 60.00 (about USD 41.00) for it.

You get the latest version of THINK Reference, but without the "Inside
Macintosh" databases. What you do get are databases for the standard
libraries C and/or C++. The TR application just lets you use the databases
provided. You still have to buy the separate product if you want the other
databases.












Path: ucivax!gateway
From: [e-mail] ("Michael D. Lyons")
Subject: PLEASE DON'T ANSWER MY QUESTION :-)
Message-ID: <[e-mail]>
X-Mailer: ELM [version 2.4 PL20]
Content-Type: text
Content-Length: 156
Newsgroups: fa.think-c
Lines: 6
Date: 28 Jul 93 08:39:20 GMT

Hi,

please don't answer my stupid THINK Ref. question, since I now have the
correct answer. My posting fingers get a little bit too itchy sometimes.

Mike










Path: ucivax!gateway
From: [e-mail] ("Michael D. Lyons")
Subject: THINK Reference & THINK C
Message-ID: <[e-mail]>
X-Mailer: ELM [version 2.4 PL20]
Content-Type: text
Content-Length: 733
Newsgroups: fa.think-c
Lines: 15
Date: 28 Jul 93 08:39:21 GMT

Hi all,

I just read in the thread about the "missing" library docs that a
copy  of THINK Reference 2.x is included in THINK C 6.0.  Is this
a full-blown version of THINK Ref, or is it a sibset of the  nor-
mal  THINK  Ref?  I'm  asking this because I was about to upgrade
from THINK C 5.0 to THINK C 6.0 *and* from THINK Ref 1.0 to 2.0.

Mike
--
+==========================================================================+
| Michael D. Lyons       | Telefon: [phone]  | EMail: [e-mail] |
| BinTec Computersysteme | Telefax: [phone] |                      |
| Willstaetter Strasse 30 / 90449 Nuernberg / Federal Republic of  GERMANY |
+==========================================================================+










Path: ucivax!gateway
From: [e-mail] (Alex Lewin)
Subject: Re: Re: No library manual?
Message-ID: <[e-mail]>
In-Reply-To: Your message of "28 Jul 93 01:56:56 GMT."
             <[e-mail]>
Newsgroups: fa.think-c
Lines: 33
Date: 28 Jul 93 12:58:15 GMT


I have Think Reference--it consists of an application and a bunch of (big!)
database files. I think the idea is that you get Think Reference minus the
Inside Mac-ish database files it usually comes with, plus the Standard
Libraries database.

I just ordered C++ 6.0, so I guess I'll find out soon!

Alex



> >From: Kathleen Fischer <[e-mail]>
> >Subject: Re:  No library manual?
> >Date: 27 Jul 93 19:40:33 GMT
> >
> >> The documentation for the standard libraries is now in a THINK Reference
> >> database. Lowers their COGS, and makes it easier to look stuff up...
> >
> >Don't you have to pay for this database?
>
> I second this question.  I have the version 5 to version 6 upgrade
> offer but reading it very carefully gives me the impression that the
> new version of Think Reference is not included.  The local distributors
> want AUD 60.00 (about USD 41.00) for it.
> --
> James Ashton                          System Administrator
> VK1XJA                         Department of Systems Engi
>neering
> Voice [phone]      Research School of Physical Sciences and Engineering
> FAX   [phone]                Australian National University
> Email [e-mail]               Canberra ACT 0200 Australia











Path: ucivax!gateway
From: [e-mail] (Steve Greenfield)
Subject: ThinkC 4.0 on Centris 650 w/System 7.1
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 22
Date: 28 Jul 93 17:04:50 GMT

When I run a simple application that uses the ANSI C library, I get the
following error:


     code segment too big (ANSI)


This must be due to the mix of Think C 4.0 and System 7.1.  Do I have
any alternatives besides buying Think C 5.something?

I got the 405-updater and have done all the updates except the scanf update.
I have run the update script but when I try and MAKE the ANSI, ANSI-A4,
ANSI-881, and ANSI-small projects I'm having trouble.  I double click on
the respective project and select MAKE from the Source menu.  I click off
Quick Scan and click Use Disk.  But when I click on MAKE it appears to go
into some type of loop.  During the compile, it keeps repeating the Lines:
and incrementing the Total:, forever!

Can someone help a frustrated user who just wants to learn Think C 4.0?


/[e-mail]










Path: ucivax!gateway
From: [e-mail]
Subject: Re: No library manual?
Message-ID: <[e-mail]>
In-reply-to: [e-mail]'s message of 27 Jul 93 19:31:04 GMT <[e-mail]>
Newsgroups: fa.think-c
Lines: 4
Date: 28 Jul 93 17:59:20 GMT

I really dislike on-line manuals.  They are hard to browse, hard to
read in bed, and hard to take on a walk in teh countryside.
Oh well, that is the way of disintegration....
==John ff










Path: ucivax!gateway
From: [e-mail] (Steve Morris)
Subject: Re: No library manual?
Message-ID: <[e-mail]>
In-Reply-To: <[e-mail]> ([e-mail])
Newsgroups: fa.think-c
Lines: 4
Date: 28 Jul 93 21:30:42 GMT


That's strange. I really dislike off-line manuals. The one I want is always
at work when I am at home or vice versa. I never remember which volume I am
after. I guess that's why both are usually available.










Path: ucivax!gateway
From: [e-mail] (Mike Greenawalt)
Subject: RE: No library manual?
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 49
Date: 29 Jul 93 02:35:30 GMT


>  From: [e-mail]
>  Subject: Re: No library manual?
>  Date: 28 Jul 93 17:59:20 GMT
>  To: [e-mail]
>
>  I really dislike on-line manuals.  They are hard to browse, hard to
>  read in bed, and hard to take on a walk in teh countryside.
>  Oh well, that is the way of disintegration....
>  ==John ff
>
>  From: Steve Morris <[e-mail]>
>  Subject: Re: No library manual?
>  Date: 28 Jul 93 21:30:42 GMT
>  To: [e-mail]
>
>
>  That's strange. I really dislike off-line manuals. The one I want is always
>  at work when I am at home or vice versa. I never remember which volume I am
>  after. I guess that's why both are usually available.
>

It appears that Symantec is going to make both forms available.  After I
started this discussion on Monday I called Symantec's customer service number
and inquired about getting the paper manual mentioned in the User's Guide.

The rep acknowledged the statement about the order form in the back of the book
was a mistake.  When asked if the manual was really available, he had to put
me on hold and go ask.  He came back and said they were not printed yet, but
would be sometime in August.  He promised that because I called Symantec would
log my interest in the paper manual and customer service would call me when
they were available.

I could not get an answer from him when I asked whether the manual would be
free or extra $$.

Anyway, I did install the upgrade last night and tried out the Think Reference
documentation for libraries.  I liked it - very nice.  But, fundamentally I
have to go along with John's thoughts (above).  I've been known to read
manuals in lots of places with no computers anywhere about.

I really appreciate all the answers I got to my question.

+---------------------------+--------------------------------------------+
| Mike Greenawalt           | [e-mail]                           |
| Metaphor Computer Systems | Metaphors be with you.                     |
| 1965 Charleston Road      |                                            |
| Mountain View, CA 94043   | These opinions are mine, not my employer's |
+---------------------------+--------------------------------------------+










Path: ucivax!gateway
From: [e-mail] (Steve Greenfield)
Subject: Dave Mark's C Primer (Vol 1)
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 7
Date: 30 Jul 93 20:33:56 GMT

Can I use Dave Mark's C Primer (Volume 1) along with Symantec's Think C
version 6.0?  Someone suggested there may be documented differences.
Thanks for any help/advice!



[e-mail]










Path: ucivax!gateway
From: [e-mail] (alan)
Subject: Explanation of CACHE problem with '040?
Message-ID: <9307302242.AA19998@mach10>
X-Mailer: LeeMail 1.2.4
Newsgroups: fa.think-c
Lines: 9
Date: 30 Jul 93 22:40:10 GMT

Can anyone provide me with a brief explanation of
why some programs execute properly with the '040
cache enabled and some fail?

Also THINKC/C++ 6.0 does not seem to offer any CACHE
control options.

Thanks.











Path: ucivax!gateway
From: [e-mail] ("Paul S. Michelsen")
Subject: Re: Explanation of CACHE problem with '040?
>From: alan <[e-mail]>
Message-ID: <[e-mail]>
In-Reply-To: <9307302242.AA19998@mach10> (message from alan on 30 Jul 93 22:40:10 GMT)
Newsgroups: fa.think-c
Lines: 25
Date: 30 Jul 93 23:51:03 GMT
>Newsgroups: fa.think-c

>
>Can anyone provide me with a brief explanation of
>why some programs execute properly with the '040
>cache enabled and some fail?
>
>Also THINKC/C++ 6.0 does not seem to offer any CACHE
>control options.
>
>Thanks.
>

One big problem is with self-modifying code.  The modified code is in
the data cache, not in the instruction cache where it needs to be.
The program thus needs to flush the code out of both caches so that
the instruction cache can load the modified code.  And this needs to
be done before the code is preloaded into instruction buffers.

There may be other problems as well.  You might want to post a summery
to the mail list when you're done.

I think there are some system or toolbox functions that do cache
flushing.  This might be in a technical note somewhere.

Good luck.











Path: ucivax!gateway
From: [e-mail] (Alan Piszcz)
Subject: Explanation of CACHE problem with '040?
Message-ID: <9307310010.AA20252@mach10>
Newsgroups: fa.think-c
Lines: 6
Date: 31 Jul 93 00:07:28 GMT


Please reply to [e-mail] or
[e-mail]

Sorry about not specifying in the previous message.











Path: ucivax!gateway
From: [e-mail] ("Mark D. Nagel")
Subject: ADMIN: let author summarize (was Re: Explanation of CACHE problem)
Message-ID: <[e-mail]>
In-Reply-To: Your message of 31 Jul 1993 00:07:28 +0000.
     <9307310010.AA20252@mach10>
Content-Type: text/plain; charset="us-ascii"
Mime-Version: 1.0
Newsgroups: fa.think-c
Reply-To: [e-mail]
Organization: Relational Data Systems, Irvine, CA
Lines: 12
Date: 31 Jul 93 03:20:40 GMT
Phone: [phone]

On 31 Jul 1993 00:07:28 GMT you wrote:

    Please reply to [e-mail] or
    [e-mail]

Please try to respond to the author and let him or her summarize to
reduce duplicate answers to the entire list.  That last round with
the Think Reference stuff was a pretty good example what can happen
otherwise...

Thanks,
Mark










Path: ucivax!gateway
From: [e-mail] (Alex Lewin)
Subject: Re:  Explanation of CACHE problem with '040?
Message-ID: <[e-mail]>
Newsgroups: fa.think-c
Lines: 32
Date: 31 Jul 93 03:36:04 GMT

> Can anyone provide me with a brief explanation of
> why some programs execute properly with the '040
> cache enabled and some fail?

I'm not sure where to start. Are you asking for a comprehensible explanation of
cache strategies and the differences among the 68020, 030 and 040 caches, or
do you understand these differences, and are curious specifically in the context
of the Mac hardware and OS?

> Also THINKC/C++ 6.0 does not seem to offer any CACHE
> control options.

There are a couple of OS calls. I'm taking this from OSUtils.h
included with C 5.x:

pascal Boolean SwapInstructionCache(Boolean cacheEnable);
pascal void FlushInstructionCache(void);
pascal Boolean SwapDataCache(Boolean cacheEnable);
pascal void FlushDataCache(void);

Actually, now that I look in THINK Reference, I see that there is a
nice little bit that would answer your first question. I'm not sure if
it would be legal for me to cut-and-post it. Anyone know? I'm looking
specifically at the entries for Flush*Cache().

If you're a real Mac programming weenie, I strongly recommend THINK
Reference. Makes interesting browsing, answers questions like yours,
and on my PowerBook, is a weight-conscious alternative to Inside Mac
and three hundred Tech Notes ;-)

Alex Lewin
Certified Mac Weenie










Path: ucivax!gateway
From: [e-mail] (Alan Piszcz)
Subject: CACHE ANSWERS
Message-ID: <9307311038.AA21761@mach10>
Newsgroups: fa.think-c
Lines: 123
Date: 31 Jul 93 10:35:52 GMT

Thank you everyone for such a quick reply.
This explains my question.

Here are the replies:

From: "Paul S. Michelsen" <[e-mail]>
Subject: Re: Explanation of CACHE problem with '040?

One big problem is with self-modifying code.  The modified code is in
the data cache, not in the instruction cache where it needs to be.
The program thus needs to flush the code out of both caches so that
the instruction cache can load the modified code.  And this needs to
be done before the code is preloaded into instruction buffers.

There may be other problems as well.  You might want to post a summery
to the mail list when you're done.

I think there are some system or toolbox functions that do cache
flushing.  This might be in a technical note somewhere.

Good luck.




From: [e-mail] (Edward J. Huff)
Subject: Re: Explanation of CACHE problem with '040?

Cache control follows the same rules as monitor depth, I believe:  only
users are supposed to change it...  or something like that.  New programs
shouldn't have trouble with cache, old programs which do have trouble
wouldn't be able to use any cache control anyway, since they were written
before the cache existed...

The 040 instruction cache is the problem, together with programs that
explicitly write into a code segment, or which contain their own private
segment loader (like programs compiled under very old versions of Think C).

A program which writes to RAM will write into the data cache, which is on
the processor chip.  A store into memory NEVER writes into the instruction
cache.  The data will also eventually be moved off chip into actual RAM, if
it is not modified again.  (I am not sure if the data cache is write
through or not).

The problem arises when a program writes into a RAM location which has
already been loaded into the instruction cache, and then branches to that
location.
The processor does NOT go out to RAM to read the location.  Instead, it
uses the contents of the instruction cache, which are incorrect.

Programs which work properly with the cache do not write into code
segments, or if they do, they arrange for the instruction cache to be
marked invalid so that instructions will be refetched.

--
Edward J. Huff   [e-mail]   [phone]
Keck Laboratory for Biomolecular Imaging
NYU Chemistry Deptartment, 31 Washington Place, New York NY 10003


From: [e-mail] (Alex Lewin)
Subject: Re:  Explanation of CACHE problem with '040?

I'm not sure where to start. Are you asking for a comprehensible explanation of
cache strategies and the differences among the 68020, 030 and 040 caches, or
do you understand these differences, and are curious specifically in the context
of the Mac hardware and OS?

There are a couple of OS calls. I'm taking this from OSUtils.h
included with C 5.x:

pascal Boolean SwapInstructionCache(Boolean cacheEnable);
pascal void FlushInstructionCache(void);
pascal Boolean SwapDataCache(Boolean cacheEnable);
pascal void FlushDataCache(void);

Actually, now that I look in THINK Reference, I see that there is a
nice little bit that would answer your first question. I'm not sure if
it would be legal for me to cut-and-post it. Anyone know? I'm looking
specifically at the entries for Flush*Cache().

If you're a real Mac programming weenie, I strongly recommend THINK
Reference. Makes interesting browsing, answers questions like yours,
and on my PowerBook, is a weight-conscious alternative to Inside Mac
and three hundred Tech Notes ;-)

Alex Lewin
Certified Mac Weenie






From: Alex Lewin <[e-mail]>
To: [e-mail]

I'm not sure where to start. Are you asking for a comprehensible explanation of
cache strategies and the differences among the 68020, 030 and 040 caches, or
do you understand these differences, and are curious specifically in the context
of the Mac hardware and OS?

There are a couple of OS calls. I'm taking this from OSUtils.h
included with C 5.x:

pascal Boolean SwapInstructionCache(Boolean cacheEnable);
pascal void FlushInstructionCache(void);
pascal Boolean SwapDataCache(Boolean cacheEnable);
pascal void FlushDataCache(void);

Actually, now that I look in THINK Reference, I see that there is a
nice little bit that would answer your first question. I'm not sure if
it would be legal for me to cut-and-post it. Anyone know? I'm looking
specifically at the entries for Flush*Cache().

If you're a real Mac programming weenie, I strongly recommend THINK
Reference. Makes interesting browsing, answers questions like yours,
and on my PowerBook, is a weight-conscious alternative to Inside Mac
and three hundred Tech Notes ;-)

Alex Lewin
Certified Mac Weenie











Path: ucivax!gateway
From: [e-mail] (Ferit M Albukrek)
Subject: Debugging Code Resources
Message-ID: <[e-mail]>
X-Mailer: ELM [version 2.3 PL11-upenn1.13]
Posted-Date: Sat, 31 Jul 93 12:07:31 EDT
Newsgroups: fa.think-c
Lines: 27
Date: 31 Jul 93 16:07:42 GMT


Hi all,

I am trying to find ways of debugging my code resources. Since these are not
stand alone applications, I am unable to use the THINK-C debugger. I tried
creating an seperate project that executes the code resource where the values
of the variables are displayed using printf statements. For some reason I
cannot get the console pop up anymore. I even tried a clean "Hello there!"
program but the console window still doesn't show up. Whenever I run the
program MacsBug complains with a bus error. It the THINK-C debugger is running,
there are no complaints but the console still doesn't show up.

I will appreciate anyone who can:

* Help me get my console window back so I can run ANSI programs.
* Suggest easy ways of debugging code resources. (specifically XCMDs) I tried
  MacsBug but it seems very complicated and I do not understand 68040 code.

Thanks a bunch!

Ferit Albukrek
[e-mail]
Language Analysis Center
University of Pennsylvania

ps: I am using THINK C 6.0 on a Mac Quadra 950.