MacTech 1 to 12/MacTech-vol-1-12.toast/Reference/the cmsp digests ('94-'97)/csmp digest Vol 4 No 019

Text File | 1996-11-18 | 60.5 KB | 1,721 lines | [TEXT/R*ch]

C.S.M.P. Digest             Sat, 09 Nov 96       Volume 4 : Issue 19

Today's Topics:

        3-D???
        An other question about Mouse move
        CopyBits performance = ü(window position)?
        Disk swapping problem - floppy reading floppy
        Documentation tool?
        Drawing to MenuBar from SystemTask patch
        FSpExchangeFiles example has a bug?
        FSpExchangeFiles-"safe-save"
        Finding memory leaks in applications using malloc, free, new, delete
        Floating windows with Text Services (NewServiceWindow)
        Flushing volumes from MacsBug
        Ftnchek 2.9.4 as an MPW tool anyone?
        GWorld problem
        GWorlds:
        Globals in PowerPC Assembly Code
        How about INITs
        How to Debug Drag & Drop ?
        How to port Think Pascal 4.0 program to PPC 604e-180?
        ICN# or ICON resource?
        Interrupt level - What am i allowed to do?
        Launch URL form FileMaker?
        Looking for a diff program.
        MacIntosh Toolbox, Def. of calls
        Meaning of error -127 ?
        Moving the mouse cursor!
        NFS client for the Macintosh
        Need program to create and customize installs for Mac
        Of Sprites and GWorlds
        Open Transport from MultiProcessor Task?
        PBCatSearch problem
        Periodic tasks w-o using VBL
        QuickTime CDEF
        Re-entrant problem?  Devices and WaitNextEvent?
        Resource Compiler
        Socket library for Macs???
        Strange code generated (CW10)
        TP and CW difference
        Those windows with thin title bars
        Tracking down stray pointers...???
        Unlocking files in AppleScript
        Visual Basic program -> Mac program
        What do you guys use for an installer of your software?
        [Q] Looping WAV files from memory.
        [Q] PBRead and the disk driver...
        [Q] SYNC PowerPC instruction not working in 604e ?
        long date-time record questions
        looking for Vantage text editor
        simple CopyBits question:



The Comp.Sys.Mac.Programmer Digest is moderated by Mark Aiken
([e-mail]).

The digest is a collection of article threads from the internet
newsgroups comp.sys.mac.programmer.help, csmp.tools, csmp.misc and
csmp.games. It is designed for people who read news semi-regularly and
want an archive of the discussions.  If you don't know what a
newsgroup is, you probably don't have access to it. Ask your systems
administrator(s) for details. If you don't have access to news, you
may still be able to post messages to the group by using a mail server
like anon.penet.fi (mail [e-mail] for more information).

Each issue of the digest contains one or more sets of articles (called
threads), with each set corresponding to a 'discussion' of a particular
subject.  The articles are not edited; all articles included in this digest
are in their original posted form (as received by our news server at
ee.mcgill.ca).  Article threads are not added to the digest until the last
article added to the thread is at least two weeks old (this is to ensure that
the thread is dead before adding it to the digest).  Article threads that
consist of only one message are generally not included in the digest.

The digests can be obtained by email, ftp or through the World Wide Web.

If you want to receive the digest by mail, send email to
[e-mail] with no subject and one of the following commands
as body:

    help                        Sends you a summary of commands
    subscribe csmp                      Adds you to the mailing list
    unsubscribe csmp                    Removes you from the list

Once you have subscribed, you will automatically receive each new
issue as it is created.

Back issues are available by ftp from Info-Mac mirror sites in the
per/csmp subdirectory, e.g.

  ftp://sumex-aim.stanford.edu/info-mac/per/csmp/

The contents of all back issues can be searched by accessing the
following URL, courtesy of Andrew Barry ([e-mail]):

    http://marvin.stattech.com.au/search.html

They can also be searched through the following URLs, thanks to
Tim Tuck ([e-mail]):

    http://wais.sensei.com.au/searchform.html
    wais://wais.sensei.com.au:210/csmp?

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

>From [e-mail]
Subject: 3-D???
Date: 24 Oct 1996 10:06:02 GMT
Organization: Deja News Usenet Posting Service

I have a great 3-D game in mind and want to start making it.  I have
made a couple of 2-D casino games and one shoot em up.   I was
wondering if I have books on quickdraw 3-D and Tricks of Mac Games
Gurus,  if I was ready to make the Jump straight into 3-D programming
?
Or is there a lot of stuff to programming in 3-D that me, as a new
programmer, am not ready for?
I know this question is broad and vague, but I don't have anyone else
to ask and I have such a good idea that I want to jump on it while
it's hot.    If all I need are books,  then I can learn it, but
does it take more than books?
Thanks
Frank La Lone
- ---------------------------------------------------------------------
This article was posted to Usenet via the Posting Service at Deja News:
http://www.dejanews.com/          [Search, Post, and Read Usenet News!]

+++++++++++++++++++++++++++

>From [e-mail] (Brian Kendall)
Date: Thu, 24 Oct 1996 18:41:11 -0400
Organization: Programmer

In article <[e-mail]>, [e-mail] wrote:

> I have a great 3-D game in mind and want to start making it.  I have
> made a couple of 2-D casino games and one shoot em up.   I was
> wondering if I have books on quickdraw 3-D and Tricks of Mac Games
> Gurus,  if I was ready to make the Jump straight into 3-D programming
> ?
> Or is there a lot of stuff to programming in 3-D that me, as a new
> programmer, am not ready for?
> I know this question is broad and vague, but I don't have anyone else
> to ask and I have such a good idea that I want to jump on it while
> it's hot.    If all I need are books,  then I can learn it, but
> does it take more than books?
> Thanks
> Frank La Lone
> -----------------------------------------------------------------------
> This article was posted to Usenet via the Posting Service at Deja News:
> http://www.dejanews.com/          [Search, Post, and Read Usenet News!]

Wow. I want to make a cool 3d game, too.
It's probably a good idea to use those two books you mentioned (if they
are a big help, lemme know :). 3d things can be a little hard, though. It
can be hard to get the perspective right. I want to make a polygon graphic
game and I have to worry about z-buffering and speed. Since there is a lot
of math and if-statements involved, it can be slowed down big time.

Someone suggested to me to just go and get some sort of package that has
ready made 3d code in it that runs right and fast. It seems like a good
idea (but where's the fun in that?)

As a final tip, I can only say one thing: don't think super duper 3d. A
macintosh (even a powermac) has it's limits. You can't have a hundred
really complex 3d objects moving around on the screen at a time unless you
have the perfect 3d code. I've lost many a good idea by trying to make
games that don't run fast enough.

Hope this helps.

Brian K.

ãããããããããããããããããããããããããããããããããããããããããããããããããããããããããããããããããã
"If you take cranberries and stew them like apple sauce, it tastes
much more like prunes then rhubarb does." ã Groucho Marx
ãããããããããããããããããããããããããããããããããããããããããããããããããããããããããããããããããã

+++++++++++++++++++++++++++

>From David Matiskella <[e-mail]>
Date: Thu, 24 Oct 1996 10:57:50 -0700
Organization: University of Washington

On 24 Oct 1996 [e-mail] wrote:

> I have a great 3-D game in mind and want to start making it.  I have
> made a couple of 2-D casino games and one shoot em up.   I was
> wondering if I have books on quickdraw 3-D and Tricks of Mac Games
> Gurus,  if I was ready to make the Jump straight into 3-D programming
> ?
> Or is there a lot of stuff to programming in 3-D that me, as a new
> programmer, am not ready for?
> I know this question is broad and vague, but I don't have anyone else
> to ask and I have such a good idea that I want to jump on it while
> it's hot.    If all I need are books,  then I can learn it, but
> does it take more than books?
> Thanks
> Frank La Lone
> -----------------------------------------------------------------------
> This article was posted to Usenet via the Posting Service at Deja News:
> http://www.dejanews.com/          [Search, Post, and Read Usenet News!]
>
YOu really should pick up a book on Computer graphics. There are basically
two types. The mathematically rigourous ones like Foley and Van Dam,
Watt's books, and a whole bunch of others, these give the theory but often
don't have much in the line of implemtation info. Then there are the 3d
game programming books. They give a lot of code but sometimes they do
weird things.

David Matiskella
[e-mail]



+++++++++++++++++++++++++++

>From [e-mail] (Bretton Wade)
Date: Fri, 25 Oct 1996 10:38:38 -0400
Organization: qualia, inc.

In article <[e-mail]>,
[e-mail] (Brian Kendall) wrote:

# game and I have to worry about z-buffering and speed. Since there is a lot

You don't necessarily need Z-buffering. Read over the BSP tree FAQ

   http://www.qualia.com/bspfaq/

then go get the demo application (windows 95/NT4.0 or MacOS)

   ftp://ftp.qualia.com/pub/bspfaq/

which has COMPLETE source code for a basic 3D renderer. This code was
intended to be demonstrative of principles, and is not built for speed. It
does not include things like texture mapping, but other publicly available
source does.

Bretton

--
Bretton Wade ([e-mail])
http://www.qualia.com/~bwade/

+++++++++++++++++++++++++++

>From [e-mail] (Kevin Stone)
Date: 25 Oct 1996 17:47:04 -0600
Organization: RapidNet

: Wow. I want to make a cool 3d game, too.
: It's probably a good idea to use those two books you mentioned (if they
: are a big help, lemme know :). 3d things can be a little hard, though. It
: can be hard to get the perspective right. I want to make a polygon graphic
: game and I have to worry about z-buffering and speed. Since there is a lot
: of math and if-statements involved, it can be slowed down big time.

   Z-buffering is not really a great idea unless you're rendering several
   thousand polygons in each frame.  Depending upon the scene that you're
   rendering, the good ol' Painters algorithm is still the best way to
   go.  If you're rendering indoor scenes or scenes where the entire
   screen is getting filled up with polygons, it might be better to use an
   inversed Painters alg... basicly you sort the polygons front-to-back
   and store the polygon's scan-line endpoints in a screen-wide
   edge-list.  The main problem is that scan-lines that overlap others will
   need to get clipped, but the final outcome is a list of scan-lines that
   can be rendered without writting any pixel more than once.

: Someone suggested to me to just go and get some sort of package that has
: ready made 3d code in it that runs right and fast. It seems like a good
: idea (but where's the fun in that?)

   Might be a really good idea for someone who dosn't have the time or
   energy to spend a year writting their own engine. ;)

: As a final tip, I can only say one thing: don't think super duper 3d. A
: macintosh (even a powermac) has it's limits. You can't have a hundred
: really complex 3d objects moving around on the screen at a time unless you
: have the perfect 3d code. I've lost many a good idea by trying to make
: games that don't run fast enough.

   My engine will do about 18,000 texture-mapped 50-pixel triangles per
   second on my 7100-66Mhz PowerMac.  In contrast, a low-end SGI will do
   about 100,000 texture-mapped triangles per second. ;)  Nintendos
   Ultra 64 will do several hundred-tousand texture-mapped tris per second.
   The new 3DO system coming out next year is boasted to do 500,000
   textured tris/second, or 1,000,000 flat color tris/second.

   I think anyone who is serious about writting 3D games is best
   off writting their own engine.  Even though it took me a year to write
   my own engine, I also wrote a fairly nice 3D modeler and other tools
   which I could spruce up and sell to the public.  I'm also planning to
   licence my engine next year after I spruce it up a bit more.

: Hope this helps.
: Brian K.

  Ditto.
  Brian S. :)

Sincerly,
  Brian Stone
  Stone Entertainment
  [e-mail]
  Check out our website at:
  http://www.rapidnet.com/~metals/stone-entertainment.html


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

>From Eric Mailloux <[e-mail]>
Subject: An other question about Mouse move
Date: Sat, 19 Oct 1996 12:04:09 -0400
Organization: Total Net

I try to use the cursor device manager for moving the cursor from a
point to an other ( thank to Samuel J. W. Leidy for is source code ) but
I cant link my code. I find a little warnig on the CursorDevice.h file
they said we cant use cursor device manager on PowerMac because it is
emulated and the mixed mode manager is wrong. Some one can tell me how
can I bypass this problem.

Sebastien

+++++++++++++++++++++++++++

>From [e-mail] (Paul Baxter)
Date: Mon, 21 Oct 1996 11:10:17 -0500
Organization: (none)

In article <[e-mail]>, Eric Mailloux <[e-mail]> wrote:

> I try to use the cursor device manager for moving the cursor from a
> point to an other ( thank to Samuel J. W. Leidy for is source code ) but
> I cant link my code. I find a little warnig on the CursorDevice.h file
> they said we cant use cursor device manager on PowerMac because it is
> emulated and the mixed mode manager is wrong. Some one can tell me how
> can I bypass this problem.
>
> Sebastien

Hi,
 You need to get the file CursorDevices.Glue.c.

--
Paul Baxter
[e-mail]

+++++++++++++++++++++++++++

>From [e-mail] (Michael Hartman)
Date: Wed, 23 Oct 1996 00:01:48 -0400
Organization: Open Cluster Software

In article <[e-mail]>, Eric Mailloux <[e-mail]> wrote:

> I try to use the cursor device manager for moving the cursor from a
> point to an other ( thank to Samuel J. W. Leidy for is source code ) but
> I cant link my code. I find a little warnig on the CursorDevice.h file
> they said we cant use cursor device manager on PowerMac because it is
> emulated and the mixed mode manager is wrong. Some one can tell me how
> can I bypass this problem.
>
> Sebastien

The latest ETO, #21, has a new version of CursorDevices.h that provides
the two routines you need for positioning the mouse.  Or you can just put
these lines in your current copy:

pascal OSErr CrsrDevMoveTo(CursorDevicePtr ourDevice, long absX, long absY)
 TWOWORDINLINE(0x7001, 0xAADB);

pascal OSErr CrsrDevNextDevice(CursorDevicePtr *ourDevice)
 TWOWORDINLINE(0x700B, 0xAADB);

Note that these are the "old" names.  These should work on all existing
Power Macs; and I suppose that when/if InterfaceLib is actually corrected
some year, these old names will be provided for compatibility.

--

Michael Hartman

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

>From [e-mail] (J.P. Djajadiningrat)
Subject: CopyBits performance = ü(window position)?
Date: 18 Oct 1996 20:41:49 GMT
Organization: Delft University of Technology

>From measuring frame rates I get the strong impression that CopyBits
performance strongly depends upon window position on the screen. If this
is so, can someone tell me the rules, or point me in the right direction?
(I've read the FAQ but couldn't find anything on window position).

Yes, bitdepth and size of offscreen GWorld and window are already the same.

Tom

+++++++++++++++++++++++++++

>From [e-mail] (Andrew Brownsword)
Date: 19 Oct 1996 00:58:04 GMT
Organization: Electronic Arts (Canada), Inc.

In article <[e-mail]>,
[e-mail] (J.P. Djajadiningrat) wrote:

>From measuring frame rates I get the strong impression that CopyBits
>performance strongly depends upon window position on the screen. If this
>is so, can someone tell me the rules, or point me in the right direction?
>(I've read the FAQ but couldn't find anything on window position).
>
>Yes, bitdepth and size of offscreen GWorld and window are already the same.
>
>Tom

To make CopyBits run faster you need to ensure that both the source and
the destination images are aligned on the machine's word boundries.  On a
680x0 it is sufficient to align both source and destination on 4 byte
boundries.  On the PowerPC there is potential benefit to 8 byte alignment,
especially if you are doing a straight copy to VRAM (no scaling, palette
conversion, depth conversion, etc).  This means that your window should be
have a horizontal position divisable by 8 in 8bit, 4 in 16bit, or 2 in
32bit.  The GWorld routines will ensure that your source is aligned on at
least 4 byte boundries (possibly 8 or 16, I'm not sure), assuming you are
copying the entire image.

--
Andrew Brownsword
Software Engineer
Electronic Arts (Canada), Inc.
- ----------------------------
"The opinions expressed in this message are my own and don't necessarily reflect those of my employer."

+++++++++++++++++++++++++++

>From David Matiskella <[e-mail]>
Date: Fri, 18 Oct 1996 18:44:40 -0700
Organization: University of Washington

On 18 Oct 1996, Gregory J Hinkle wrote:

> J.P. Djajadiningrat ([e-mail]) wrote:
> : From measuring frame rates I get the strong impression that CopyBits
> : performance strongly depends upon window position on the screen. If this
> : is so, can someone tell me the rules, or point me in the right direction?
> : (I've read the FAQ but couldn't find anything on window position).
>
> : Yes, bitdepth and size of offscreen GWorld and window are already the same.
>
> : Tom
>
> I believe it is fastest if the origin of the pixmap in memory is
> divisible by 4. Or X by Y is divisible by four.
>
I think develop 11 had an article on it. The article is a bit dated as
it's pre powermac. That being said On a power mac you want to align to
doubles if  you are going to be using doubles for copying. Your y location
shouldn't matter.  That being said I believe NewGWorld returns a 4 byte
aligned pixelmap. The solution I have seen is to create a pixmap bigger
than the one you need and just ignore the first couple columns which are
not aligned. I am sure that this something that will (maybe already has)
change in the future.

David Matiskella
[e-mail]



+++++++++++++++++++++++++++

>From [e-mail] (Scott Squires)
Date: Fri, 18 Oct 1996 22:25:02 -0800
Organization: Puffin Designs

In article <[e-mail]>,
[e-mail] (Andrew Brownsword) wrote:

>In article <[e-mail]>,
>[e-mail] (J.P. Djajadiningrat) wrote:
>
>>From measuring frame rates I get the strong impression that CopyBits
>>performance strongly depends upon window position on the screen. If this
>>is so, can someone tell me the rules, or point me in the right direction?
>>(I've read the FAQ but couldn't find anything on window position).
>>
>>Yes, bitdepth and size of offscreen GWorld and window are already the same.
>>
>>Tom
>
>To make CopyBits run faster you need to ensure that both the source and
>the destination images are aligned on the machine's word boundries.

You can use routines in QuickDraw to AlignWindows so they snap to
the correct placement.  You can also do an UpdateGWorld and pass it
global coordniates.  It will use the extra bytes of the GWorld rowbytes
to align it to the window.

-scott


Scott Squires               "Insert funny stuff here"
[e-mail]
[e-mail]


+++++++++++++++++++++++++++

>From [e-mail] (Gregory J Hinkle)
Date: 18 Oct 1996 22:17:47 GMT
Organization: University of Pennsylvania

J.P. Djajadiningrat ([e-mail]) wrote:
: From measuring frame rates I get the strong impression that CopyBits
: performance strongly depends upon window position on the screen. If this
: is so, can someone tell me the rules, or point me in the right direction?
: (I've read the FAQ but couldn't find anything on window position).

: Yes, bitdepth and size of offscreen GWorld and window are already the same.

: Tom

I believe it is fastest if the origin of the pixmap in memory is
divisible by 4. Or X by Y is divisible by four.

--
*******************************************************************************
*   _/_/_/    _/    _/  |  University of Pennsylvania, School of Engineering  *
*  _/    _/   _/    _/  |---------------------------------------------------- *
*  _/         _/    _/  |  Box 0859                  |  Drive Defensivly...   *
*  _/  _/_/   _/_/_/_/  |  3901 Locust Walk          |      Buy a Tank.       *
*  _/    _/   _/    _/  |  Philadelphia, Pa 19104    |    --Greg Hinkle--     *
*    _/_/     _/    _/  |-----------------------------------------------------*
*_______________________|        http://homepage.seas.upenn.edu/~ghinkle/     *
*    Greg      Hinkle   |             [e-mail]            *
*******************************************************************************

+++++++++++++++++++++++++++

>From [e-mail] (J.P. Djajadiningrat)
Date: 23 Oct 1996 09:03:48 GMT
Organization: Delft University of Technology

Many thanks to all who replied to my question!

Tom

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

>From "Brian Mowatt" <[e-mail]>
Subject: Disk swapping problem - floppy reading floppy
Date: 17 Oct 1996 10:28:30 GMT
Organization: Crocodile Clips Ltd.

 Hi.
My problem is this I wish to run a program from a floppy that requires the
user to put other floppies in. Because its part of an installation process
I cant really copy the program to the hard disk and run it from there.
 I'm getting a ridiculous amount of disk swaps - its a REAL pain - and
unusuable. Im no expert on the mac memory management, but I reckon the mac
is needing the original (program containing) disk every other line of code,
in order to know what to do. The solution would probably be to preload ALL
the code and resource segments i need to use, and lock them somehow - but i
cant seem to find out how to do this...I can find plenty about locking say
some data you've put on the heap etc.
Does anyone know a simple way round this. Something like
"look here system, load ALL of me, and don't unload me untill i'm finished.
Just leave me ALONE!"
my prog. isn't big, and not run for long, so such system unfriendly
behaviour shouldn't be a problem.
There must be loads of people who've encountered this problem!
Thanks in advance.
--
[e-mail]
Software Developer
Crocodile Clips ltd.


+++++++++++++++++++++++++++

>From [e-mail] (Jump Long)
Date: 18 Oct 1996 00:04:27 -0400
Organization: America Online, Inc. (1-800-827-6364)

Brian Mowatt wrote:

> Hi.  My problem is this I wish to run a program from a floppy
>that requires the user to put other floppies in. Because its
>part of an installation process I cant really copy the program
>to the hard disk and run it from there.
> I'm getting a ridiculous amount of disk swaps - its a REAL pain
>- and unusuable. Im no expert on the mac memory management, but
>I reckon the mac is needing the original (program containing)
>disk every other line of code, in order to know what to do. The
>solution would probably be to preload ALL the code and resource
>segments i need to use, and lock them somehow - but i cant seem
>to find out how to do this...I can find plenty about locking say
>some data you've put on the heap etc. Does anyone know a simple
>way round this. Something like  "look here system, load ALL of
>me, and don't unload me untill i'm finished. Just leave me
>ALONE!" my prog. isn't big, and not run for long, so such system
>unfriendly behaviour shouldn't be a problem. There must be loads
>of people who've encountered this problem! Thanks in advance.

Make your resources preload and non-purgable and you'll probably solve
your problems.

If after that, you're *still* getting the Disk Switch dialog, you can
break at the address DskSwtchHook (3EA) points to with this MacsBug
command:

  BR 3EA^ ';DM RA0 iopb'

That will break into MacsBug, right before the File Manager's Disk Switch
code is run, with the parameter block pointer of the call causing the I/O
to the offline/ejected disk in register A0. The ';DM RA0 iopb' part of
that command line tells MacsBug to show you the parameter block.

>From there, you should be able to look at the ioTrap field to see what
File Manager call is trying to execute and what it is trying to access on
the offline/ejected disk. ioTrap will either have the A-Trap work in it,
or if the high nibble isn't "A", then it will have the HFSDispatch trap
selector in the low byte.

- Jim Luther

+++++++++++++++++++++++++++

>From "Brian Mowatt" <[e-mail]>
Date: 18 Oct 1996 15:09:19 GMT
Organization: Crocodile Clips Ltd.

> I wrote:
>
> > Hi.  My problem is this I wish to run a program from a floppy
> >that requires the user to put other floppies in. Because its
> >part of an installation process I cant really copy the program
> >to the hard disk and run it from there.
> > I'm getting a ridiculous amount of disk swaps - its a REAL pain
> >- and unusuable.  .. ...
etc.
.........There must be loads
> >of people who've encountered this problem! Can anyone help?
  Thanks in advance.

Jump Long <[e-mail]> replied in  in article
<546vkb$[e-mail]>...
>

> Make your resources preload and non-purgable and you'll probably solve
> your problems.
>
> If after that, you're *still* getting the Disk Switch dialog, you can
> break at the address DskSwtchHook (3EA) points to with this MacsBug
> command:
>
>   BR 3EA^ ';DM RA0 iopb'  ..

 .. and a lot more besides

- Jim Luther

 Thanks Jim, but unfortunately that doesn't help much. I went through my
code segements with resedit making them preload no purge, (and
interestingly i HAD to lock some when i did this or things fell over! -
strange that it was exactly the same code that normally ran OK, just
resource attributes altered in resedit..)
 Unfortunately what I know about macsbug could be written on the edge of a
postage stamp, so the more low level stuff is a bit (though not
completely!) lost on me. I dont have the time to learn about macsbug do
you.. or anyone else have and more suggestions?

 I though if i preloaded and locked everything all would be OK. But it
wasnt, and i'm fast running out of ideas!
Please .. someone!!

[e-mail]
Software developer.
Crocodile clips ltd.

+++++++++++++++++++++++++++

>From [e-mail] (Jump Long)
Date: 24 Oct 1996 01:26:21 -0400
Organization: America Online, Inc. (1-800-827-6364)

Brian Mowatt wrote:
>Unfortunately what I know about macsbug could be written on the
>edge of a postage stamp, so the more low level stuff is a bit
>(though not completely!) lost on me. I dont have the time to
>learn about macsbug do you.. or anyone else have and more
>suggestions?

Take the time to learn MacsBug. It's my favorite debugging tool and with
it, you'll be able to do things high-level debuggers just dream of.

Jim Luther
MacsBug is the source level debugger for my assembler

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

>From [e-mail] (Mikael Skogström)
Subject: Documentation tool?
Date: 17 Oct 1996 10:04:25 GMT
Organization: TechTalk

Is there a Mac based documentation tool to ease the documentation
process associated with software development? For example, an
application that can interprete C++ and Pascal code, and the use of
different classes, and turn it into some kind of a flow chart? And
maybe even display the gathered the information in a number of other
ways?
 If there isn't such a tool for the Mac, maybe you know of one for PCs?

Mikael

+++++++++++++++++++++++++++

>From [e-mail] (Mikael Skogström)
Date: 19 Oct 1996 09:54:59 GMT
Organization: TechTalk

In article <jordanz-1710960922570001@204.147.232.52>
[e-mail] (Jordan Zimmerman) writes:

> Object Master is good at this.

Okay. Who's manufacturing it and do I find them on the Net?

Mikael

+++++++++++++++++++++++++++

>From [e-mail] (Jordan Zimmerman)
Date: Thu, 17 Oct 1996 09:22:26 -0800
Organization: Altura Software, Inc.

In article <5450b9$[e-mail]>, [e-mail] (Mikael
Skogström) wrote:

> Is there a Mac based documentation tool to ease the documentation
> process associated with software development? For example, an
> application that can interprete C++ and Pascal code, and the use of
> different classes, and turn it into some kind of a flow chart? And
> maybe even display the gathered the information in a number of other
> ways?
>  If there isn't such a tool for the Mac, maybe you know of one for PCs?
>
> Mikael

Object Master is good at this.

--
Jordan Zimmerman
Altura Software, Inc.
home page: http://www.altura.com/jordanz

Harry Browne for President!
http://www.harrybrowne96.org   1 [phone]
Stop the Browne Out! Let Harry Browne debate!     http://www.twr.com/STBO

+++++++++++++++++++++++++++

>From [e-mail] (David Rehring)
Date: Wed, 23 Oct 1996 17:48:45 -0700
Organization: GDT Softworks, Inc.

In article <54a8hj$[e-mail]>, [e-mail] (Mikael
Skogström) wrote:

>>In article <jordanz-1710960922570001@204.147.232.52>
>>[e-mail] (Jordan Zimmerman) writes:
>>
>>> Object Master is good at this.
>>
>>Okay. Who's manufacturing it and do I find them on the Net?
>>
>>Mikael

ACI US

try www.acius.com

--
David Rehring
Senior Software Engineer
GDT Softworks, Inc.
And all around insane guy!

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

>From Moritz Borgmann <[e-mail]>
Subject: Drawing to MenuBar from SystemTask patch
Date: Fri, 25 Oct 1996 22:56:41 +0100
Organization: none

I guess this is something for the real mac programming gurus :-)

What I am trying to accomplish is the following: I want to display
some clock-like information in the menu bar, exactly as system 7.5
does. This should be updated in real time and look nice, without
flickering.

So I wrote an INIT which installs two trap patches: DrawMenuBar (tail
patch) and SystemTask (head patch). The first time the DrawMenuBar
patch is called, it allocates a new offscreen GWorld via NewGWorld in
the System heap.
Apart from that, both patches do exactly the same:

-SetZone(SystemZone()), so that any new data structures will be
allocated in the system heap and not in the current heap which may be
any application's heap
-allocate space for an new CGrafPort via NewPtr
-open the new CGrafPort via OpenCPort (drawing to this port draws
directly to the desktop)
-if necessary, call UpdateGWorld to reflect changes in the size of the
stuff to be displayed
-get the current GWorld
-SetGWorld to the offscreen world
-draw the stuff
-set the GWorld back
-copy the offscreen data into the menu bar area of the new port using
CopyBits
-close the new port and dispose of the memory
-set the current zone back

Most of the time, this works fine. Everything is displayed very
smoothly, without flicker, the system isn't slowed down. But very
seldomly, my Mac crashes, and I suspect that my routines are causing
the trouble. Thus, my question is: Is it safe to do what I'm doing? Am
I missing something that I should pay attention to? Could I do this in
a less ugly way (without trap patching)?

Any advice will be greatly appreciated,

Moritz

+++++++++++++++++++++++++++

>From Moritz Borgmann <[e-mail]>
Date: Sat, 26 Oct 1996 16:24:38 +0100
Organization: none

Tom Dowdy wrote:
>you do seem to be missing a set/restore port/device around the open/close
>c ports.
>
>Some apps may assume that the current port doesn't change across a call
>to SystemTask().

Eh, sorry, I forgot to tell you that I of course DO save/restore the
port. Anything else would be crashing the mac right away, I guess.

Scott Kevill wrote:
>Are you calling LockPixels() on the GWorld PixMapHandle before drawing
>into the GWorld?

Yeah, I do that, too.
This is what makes me a bit desperate: I seem to obeye all the rules of
dealing with offscreen graphics, but it still crashes sometimes. Thanks
anyway for your replies.

Moritz
[e-mail]

+++++++++++++++++++++++++++

>From [e-mail] (Scott Kevill)
Date: Sat, 26 Oct 1996 11:23:46 +0800
Organization: Curtin University of Technology

In article <[e-mail]>,
[e-mail] wrote:

:Most of the time, this works fine. Everything is displayed very
:smoothly, without flicker, the system isn't slowed down. But very
:seldomly, my Mac crashes, and I suspect that my routines are causing
:the trouble. Thus, my question is: Is it safe to do what I'm doing? Am
:I missing something that I should pay attention to? Could I do this in
:a less ugly way (without trap patching)?
:
:Any advice will be greatly appreciated,

Are you calling LockPixels() on the GWorld PixMapHandle before drawing
into the GWorld?

That's all that comes to mind at the moment.

Hope this helps.

Scott Kevill
[e-mail]

:
:Moritz

+++++++++++++++++++++++++++

>From [e-mail] (Chris Kelly)
Date: Sat, 26 Oct 1996 12:02:05 -0800
Organization: Samizdat Productions

Have you tried doing your memory allocations just once, instead of each
time your patch is called?

Hope that helps.

Chris Kelly

In article <[e-mail]>,
[e-mail] wrote:

> Tom Dowdy wrote:
> >you do seem to be missing a set/restore port/device around the open/close
> >c ports.
> >
> >Some apps may assume that the current port doesn't change across a call
> >to SystemTask().
>
> Eh, sorry, I forgot to tell you that I of course DO save/restore the
> port. Anything else would be crashing the mac right away, I guess.
>
> Scott Kevill wrote:
> >Are you calling LockPixels() on the GWorld PixMapHandle before drawing
> >into the GWorld?
>
> Yeah, I do that, too.
> This is what makes me a bit desperate: I seem to obeye all the rules of
> dealing with offscreen graphics, but it still crashes sometimes. Thanks
> anyway for your replies.
>
> Moritz
> [e-mail]

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

>From [e-mail] (Kevin Ashbridge)
Subject: FSpExchangeFiles example has a bug?
Date: Sun, 20 Oct 1996 19:58:57 +0200
Organization: Rainy Night Software

SmartFriends,

I'm working my way through IM:Files. In the example given to "safe save" a
document, the code snippet (Listing 1-10; shown below) returns an fBsyErr
in the final statement when compiled. I don't understand why it should. The
fBsyErr error occurs when FSpDelete is called, indicating "File busy,
directory not empty, or working directory control block open". However, the
code snippet has explicitly called FSClose. What gives?

-- Kevin.


  myFSpec := myData^^.fileFSSpec;           {get FSSpec for existing file}

  GetDateTime(myTime);                      {create a temporary filename}
  NumToString(myTime, myName);

  {Find the temporary folder on file's volume; create it if necessary.}
  myErr := FindFolder(myFSpec.vRefNum, kTemporaryFolderType,
                       kCreateFolder, myVRef, myDirID);
  IF myErr = noErr THEN                     {make an FSSpec for temp file}
     myErr := FSMakeFSSpec(myVRef, myDirID, myName, myTSpec);
  IF (myErr = noErr) OR (myErr = fnfErr) THEN{create a temporary file}
     myErr := FSpCreate(myTSpec, 'trsh', 'trsh', smSystemScript);
  IF myErr = noErr THEN                     {open the newly created file}
     myErr := FSpOpenDF(myTSpec, fsRdWrPerm, myTemp);
  IF myErr = noErr THEN                     {write data to the data fork}
     myErr := DoWriteData(myWindow, myTemp);
  IF myErr = noErr THEN                     {close the temporary file}
     myErr := FSClose(myTemp);
  IF myErr = noErr THEN                     {swap data in the two files}
     myErr := FSpExchangeFiles(myTSpec, myFSpec);
  IF myErr = noErr THEN                     {delete the temporary file}
     myErr := FSpDelete(myTSpec);  <--- fBsyErr here!!




+++++++++++++++++++++++++++

>From "K. J. Bricknell" <[e-mail]>
Date: Mon, 21 Oct 1996 17:57:48 +1100
Organization: Spirit Networks - Canberra, Australia

Kevin Ashbridge wrote:
>
> I'm working my way through IM:Files. In the example given to "safe save" a
> document, the code snippet (Listing 1-10; shown below) returns an fBsyErr
> in the final statement when compiled ...

Kevin,

That one drove me bonkers for a while, too.  The only way I could find
around it was as as per the following example:

*****************************************************************************

OSErr  doWriteFile(WindowPtr windowPtr,Boolean newFile)
{
  docRecordHandle  docRecHdl;
  FSSpec           fileSpecActual, fileSpecTemp;
  UInt32           currentTime;
  Str255           tempFileName;
  SInt16           tempFileVolNum, tempFileRefNum;
  SInt32           tempFileDirID;
  OSErr            osError;

  docRecHdl = (docRecordHandle) GetWRefCon(windowPtr);
  fileSpecActual = (*docRecHdl)->fileFSSpec;

  GetDateTime(¤tTime);
  NumToString((SInt32) currentTime,tempFileName);

  osError = FindFolder(fileSpecActual.vRefNum,kTemporaryFolderType,
                       kCreateFolder &tempFileVolNum,&tempFileDirID);
  if(osError == noErr)
    osError = FSMakeFSSpec(tempFileVolNum,tempFileDirID,tempFileName,
                           &fileSpecTemp);
  if(osError == noErr || osError == fnfErr)
    osError = FSpCreate(&fileSpecTemp,'trsh','trsh',smSystemScript);
  if(osError == noErr)
    osError = FSpOpenDF(&fileSpecTemp,fsRdWrPerm,&tempFileRefNum);
  if(osError == noErr)
    osError = doWriteTextData(windowPtr,tempFileRefNum);
  if(osError == noErr)
    osError = FSClose(tempFileRefNum);
  if(osError == noErr)
    osError = FSClose((*docRecHdl)->fileRefNum);// NEW: CLOSE THE ACTUAL
FILE
  if(osError == noErr)
    osError = FSpExchangeFiles(&fileSpecTemp,&fileSpecActual);
  if(osError == noErr)
    osError = FSpDelete(&fileSpecTemp);
  if(osError == noErr)                          // THEN OPEN IT AGAIN!
    osError =
FSpOpenDF(&fileSpecActual,fsRdWrPerm,&(*docRecHdl)->fileRefNum);

  if(osError == noErr)
  {
    if(newFile)
      osError = doCopyAppNameResource(windowPtr);
  }

  return(osError);
}

******************************************************************************

This has all the logic of finding an open door and then closing it
before opening it again to walk through it.  However, it is the only way
I could get FSpExchangeFiles to behave itself.

I have seen other writings which indicate that there is something amiss
with either FSpExchangeFiles or the way it is documented.  Given your
experience, I am now convinced that that is the case.

K. J. Bricknell
Canberra
Australia

+++++++++++++++++++++++++++

>From [e-mail] (Jump Long)
Date: 24 Oct 1996 01:51:21 -0400
Organization: America Online, Inc. (1-800-827-6364)

Kevin Ashbridge wrote:

>I'm working my way through IM:Files. In the example given to
>"safe save" a document, the code snippet (Listing 1-10; shown
>below) returns an fBsyErr in the final statement when compiled.
>I don't understand why it should. The fBsyErr error occurs when
>FSpDelete is called, indicating "File busy, directory not empty,
>or working directory control block open". However, the code
>snippet has explicitly called FSClose. What gives?

The code snippet is wrong. For an good explanation of what
FSpExchangeFiles and PBExchangeFiles really do, see the Macintosh
Technical Note "Inside Macintosh Files Errata".

- Jim Luther

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

>From Thomas Wurst <[e-mail]>
Subject: FSpExchangeFiles-"safe-save"
Date: Tue, 22 Oct 1996 20:15:12 +0200
Organization: InterNetNews at ZDV Uni-Tuebingen

Hello all!

To save my documents in a safe way I'm using the "safe-save" mechanism
described in IM Files (create a temporary file in the trash using the
number of ticks since startup as filename, write all the data to that file,
then exchange the document file with the temporary one).
But after doing so, and then restarting my Mac, I suddenly have all the
created temporary files (now empty, of course) lurking in my trash!

What I'm wondering right now is this: does the "safe-save" mechanism from
IM Files all the work for me (including the proper deletion of all the
temporary files) -- would mean I'm making something wrong in my
implementation of that example code -- or do I have to delete them on my
own?

Any comments on this appreciating: Thomas.

+++++++++++++++++++++++++++

>From [e-mail] (Jump Long)
Date: 24 Oct 1996 01:57:31 -0400
Organization: America Online, Inc. (1-800-827-6364)

>To save my documents in a safe way I'm using the "safe-save"
>mechanism  described in IM Files (create a temporary file in the
>trash using the  number of ticks since startup as filename,
>write all the data to that file,  then exchange the document
>file with the temporary one). But after doing so, and then
>restarting my Mac, I suddenly have all the  created temporary
>files (now empty, of course) lurking in my trash!
>
>What I'm wondering right now is this: does the "safe-save"
>mechanism from  IM Files all the work for me (including the
>proper deletion of all the  temporary files) -- would mean I'm
>making something wrong in my  implementation of that example
>code -- or do I have to delete them on my  own?

The code snippet in Inside Macintosh: Files showing how to use
ExchangeFiles is wrong. See the Macintosh Technical Note "Inside
Macintosh: Files - Errata" for an explanation of how the ExchangeFiles
routines really work.

- Jim Luther

+++++++++++++++++++++++++++

>From [e-mail] (Eric Dorland)
Date: Thu, 24 Oct 1996 16:28:02 -0400
Organization: Odyssee Internet

In article
<[e-mail]>,
Thomas Wurst <[e-mail]> wrote:

> Hello all!
>
> To save my documents in a safe way I'm using the "safe-save" mechanism
> described in IM Files (create a temporary file in the trash using the
> number of ticks since startup as filename, write all the data to that file,
> then exchange the document file with the temporary one).
> But after doing so, and then restarting my Mac, I suddenly have all the
> created temporary files (now empty, of course) lurking in my trash!
>
> What I'm wondering right now is this: does the "safe-save" mechanism from
> IM Files all the work for me (including the proper deletion of all the
> temporary files) -- would mean I'm making something wrong in my
> implementation of that example code -- or do I have to delete them on my
> own?

You have to make sure you delete the temporary file in the code using
FSpDelete when your done. Check page 1-25, 1-26 in IM: Files for example
code on how to do it. Right at the end you'll see the FSpDelete.

Eric Dorland
[e-mail]
http://www.odyssee.net/~hooty/

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

>From [e-mail] (Frank Kim)
Subject: Finding memory leaks in applications using malloc, free, new, delete
Date: 22 Oct 1996 14:08:12 GMT
Organization: MIT Laboratory for Computer Science, Cambridge MA

Hi,

What do you guys recommend using for finding memory leaks in
applications using malloc, free, new, delete?

I've been using ZoneRanger but it doesn't seem to work too well.  I am
hoping to find something like BoundsChecker on the PC.

By the way, ever notice that when you run a program with SIOUX console,
that the program just freezes when it runs out of memory and you have to
reboot.  Seems like a shortcoming in the way SIOUX was programmed, don't
you think?
--
Frank Kim    [e-mail]
        http://cag-www.lcs.mit.edu/~frankkim/


+++++++++++++++++++++++++++

>From [e-mail] (MW Ron)
Date: Tue, 22 Oct 1996 12:19:36 -0400
Organization: Metrowerks

In article <54ikgc$[e-mail]>,
[e-mail] (Frank Kim) wrote:

>Hi,
>
>What do you guys recommend using for finding memory leaks in
>applications using malloc, free, new, delete?

Using DebugNew is the way to go.

>By the way, ever notice that when you run a program with SIOUX console,
>that the program just freezes when it runs out of memory and you have to
>reboot.  Seems like a shortcoming in the way SIOUX was programmed, don't
>you think?

I have had reports that if you run out of memory before you have written
any output to SIOUX it will hang when you try to write the error message
to the SIOUX window because there is not enough memory to initialize the
managers. If this is the error you mean, I am not sure if this is a
limitation, bug, or considered user error.

Ron

--
METROWERKS                   Ron Liechty
"Software at Work"    [e-mail]
http://www.metrowerks.com/about/people/rogues.html#mwron

+++++++++++++++++++++++++++

>From [e-mail] (Andrew Wulf)
Date: Tue, 22 Oct 1996 15:59:11 -0500
Organization: Biit, Inc.

In article <54ikgc$[e-mail]>,
[e-mail] (Frank Kim) wrote:

>Hi,
>
>What do you guys recommend using for finding memory leaks in
>applications using malloc, free, new, delete?
>
>I've been using ZoneRanger but it doesn't seem to work too well.

(1) You can use Metrowerks supplied DebugNew.

(2) For operator new/delete you can edit the file new.cp in the MWCRuntime
library, and change its setting to use NewPtr/DeletePtr. This way you can
use ZoneRanger.

>I am hoping to find something like BoundsChecker on the PC.

(3) Soon you can use my product, HeapManager, which includes an allocator
library with a debugging API, and an external application that displays
(among it's other features) all current memory allocations that came from
new/malloc, with the routine name that generated them, the size, the
address, and the order they were created. It also has a history window
that shows all allocations and frees in order with the same info. It will
be shipping in December. See the web page below for more info.

--
Andrew Wulf
Biit, Inc.



+++++++++++++++++++++++++++

>From [e-mail] (Giuliano Gavazzi)
Date: Thu, 24 Oct 1996 21:13:25 +0100
Organization: myself

In article <[e-mail]>, [e-mail]
(Andrew Wulf) wrote:

...
> be shipping in December. See the web page below for more info.
>
> --
> Andrew Wulf
> Biit, Inc.

below what?


Giuliano

+++++++++++++++++++++++++++

>From [e-mail] (Andrew Wulf)
Date: Fri, 25 Oct 1996 09:49:52 -0500
Organization: Biit, Inc.

In article <[e-mail]>,
[e-mail] (Giuliano Gavazzi) wrote:

>...
>> be shipping in December. See the web page below for more info.
>>
>> --
>> Andrew Wulf
>> Biit, Inc.
>
>below what?

Sorry, wrong signature. Try this one:

--
Andrew Wulf
Biit, Inc.
See info on HeapManager at <http://www.biit.com>

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

>From [e-mail] (Gisli R. Hjaltason)
Subject: Floating windows with Text Services (NewServiceWindow)
Date: 22 Oct 1996 13:37:06 -0400
Organization: Dept. of Computer Science, University of Maryland, College Park

I am adding a tool bar to an app that I'm working on. It seemed like the
floating window service in the Text Services (NewServiceWindow etc.) fit the
bill for making the tool bar float above the regular app windows. I've
gotten it to work, but there is one slight problem: if my app crashes, the
tool bar stays on the screen. Is there any way around this? (Of course, I'd
like to say that my app is bug-free, but we all know how these pesky bugs
can lurk around.)

I'd appreciate any advice.

Thanks in advance...
--
Gisli Runar Hjaltason              \  "Once, in the army, I was asked if I was
Department of Computer Science      \  really the composer Arnold Schoenberg.
University of Maryland, College Park \ 'Somebody had to be,' I said" Schoenberg

+++++++++++++++++++++++++++

>From [e-mail] (Brian Clark)
Date: Tue, 22 Oct 1996 15:06:33 -0500
Organization: Fawlty Software Ltd.

In article <54j0o2$[e-mail]>, [e-mail] (Gisli
R. Hjaltason) wrote:

>I am adding a tool bar to an app that I'm working on. It seemed like the
>floating window service in the Text Services (NewServiceWindow etc.) fit the
>bill for making the tool bar float above the regular app windows. I've
>gotten it to work, but there is one slight problem: if my app crashes, the
>tool bar stays on the screen. Is there any way around this?

My suggestion would be to skip the Text Services floater and use the
floating windows package from develop magazine. It does not suffer from the
drawbacks (no color support, can't change cursor when it's over the
floating window, unusual event handling) that the Text Services floaters
do. The code from develop is a bit more work to use, but worth it.

--
Thanks to everyone who has registered YA-NewsWatcher.
For the latest version (2.3.1) and helper apps, go to:
ftp://ftp.enteract.com/users/baclark/

+++++++++++++++++++++++++++

>From Jae Ho Chang <[e-mail]>
Date: Wed, 23 Oct 1996 21:03:11 +0200
Organization: eMusicas softwares

Gisli R. Hjaltason wrote:

> I've gotten it to work, but there is one slight problem: if my app crashes, the
> tool bar stays on the screen. Is there any way around this?

This could work well:
Write a jGNEFilter for the window and put it in the system heap. The
filter procedure is alive even if your program crashes.

Jae Ho Chang :-)
--
Check out the amazing note utility FinderNote at
<http://www.xs4all.nl/~jaeho/FinderNote>
eMusicas Home Page <http://www.xs4all.nl/~jaeho/>

- Let's think of the earth -


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

>From [e-mail] (Michael Schuerig)
Subject: Flushing volumes from MacsBug
Date: Sun, 20 Oct 1996 01:52:24 +0200
Organization: RHRZ - University of Bonn (Germany)


Is there a way to flush volumes from inside MacsBug? My intention, of
course, is to write the cache contents to disk after crashing into
MacsBug, but I don't necessarily want to restart (with rs).

Michael

- -
Michael Schuerig
mailto:[e-mail]
http://www.rhrz.uni-bonn.de/~uzs90z/

+++++++++++++++++++++++++++

>From [e-mail] (AndrewWelc)
Date: 21 Oct 1996 07:16:44 -0400
Organization: America Online, Inc. (1-800-827-6364)

> Is there a way to flush volumes from inside MacsBug? My intention, of
> course, is to write the cache contents to disk after crashing into
> MacsBug, but I don't necessarily want to restart (with rs).

I suppose you could always write a 'dcmd' that attempts to do this.

+--------------------------------------------------------------+
|    Andrew Welch - Thaumaturgist - Ambrosia Software, Inc.    |
+-------------------------------+------------------------------+
|    AOL-> Keyword: Ambrosia    |  http://www.AmbrosiaSW.com/  |
|    CIS-> GO word: Ambrosia    |   ftp://ftp.AmbrosiaSW.com/  |
+-------------------------------+------------------------------+

+++++++++++++++++++++++++++

>From Jim Murphy <[e-mail]>
Date: 21 Oct 1996 22:16:59 GMT
Organization: Apple Computer, Inc.

In article <[e-mail]>
Michael Schuerig, [e-mail] writes:
>Is there a way to flush volumes from inside MacsBug? My intention, of
>course, is to write the cache contents to disk after crashing into
>MacsBug, but I don't necessarily want to restart (with rs).

    If you're using MacsBug 6.5.3 (you should be), try logging something
using the log command. A sequence like this should do the trick:

    LOG logvol:dummylogfile; LOG

   (Where "logvol" is the volume you want to flush.)

   MacsBug 6.5.3 is extremely paranoid about getting what you logged onto
disk. If it's safe to do so (the file system isn't busy), it will do a
FlushVol after each line of output is drawn, and also when it closes the
file. You obviously need to be aware that if the crash in question somehow
wrote data over the file system cache, there is a chance you could be
writing bad data back to the disk. Buyer beware.


Jim Murphy                  Dr. MacsBug                    [e-mail]

"...never quite said what I wanted to say to you,                 The Cure
  never quite managed the words to explain to you..."             Untitled

+++++++++++++++++++++++++++

>From [e-mail] (Michael Schuerig)
Date: Mon, 21 Oct 1996 22:51:16 +0200
Organization: RHRZ - University of Bonn (Germany)

AndrewWelc <[e-mail]> wrote:

> > Is there a way to flush volumes from inside MacsBug? My intention, of
> > course, is to write the cache contents to disk after crashing into
> > MacsBug, but I don't necessarily want to restart (with rs).
>
> I suppose you could always write a 'dcmd' that attempts to do this.

So then, does anybody have a dcmd shell for CodeWarrior available?

Michael

- -
Michael Schuerig
mailto:[e-mail]
http://www.rhrz.uni-bonn.de/~uzs90z/

+++++++++++++++++++++++++++

>From [e-mail] (Jump Long)
Date: 24 Oct 1996 01:48:05 -0400
Organization: America Online, Inc. (1-800-827-6364)

Andrew Welch  replied:

>> Is there a way to flush volumes from inside MacsBug? My intention, of
>> course, is to write the cache contents to disk after crashing into
>> MacsBug, but I don't necessarily want to restart (with rs).
>
>I suppose you could always write a 'dcmd' that attempts to do this.

Just remember that because the File Manager may be busy and MacsBug has
the system stopped, attempting to flush a volume from within MacsBug may
just lock up your system completely.

- Jim Luther

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

>From [e-mail] (Mark Wardle)
Subject: Ftnchek 2.9.4 as an MPW tool anyone?
Date: 18 Oct 1996 05:54:00 GMT
Organization: RCfTA, U. Sydney

In article <[e-mail]>,
[e-mail] (Dr. Stefan P. Mueller) wrote:

>
> We would like to have a Macintosh version of ftnchek 2.9.4 (preferably a
> fat binary). The distribution includes a makefile for the Macintosh using
> MPW C. Unfortunately, we don't have this compiler available here. Has
> anybody with MPW C built the new version and would be willing to make the
> Mac version available?
>


Any chance of somebody making an updated MPW tool (68k) available?
The version at switch is 2.7, I believe.

Are there any other fortran77-related MPW tools I should know about?

Thanks!

+++++++++++++++++++++++++++

>From [e-mail] (M.Evrard)
Date: Mon, 21 Oct 1996 11:25:09 +0100
Organization: Royal Military Academy (ERM/KMS), Brussels

In article <[e-mail]>,
[e-mail] (Mark Wardle) wrote:

>
> Any chance of somebody making an updated MPW tool (68k) available?
> The version at switch is 2.7, I believe.
>

I uploaded this weekend the MPW tool version to R.Moniot (the author of ftnchek)
and it should appear very soon at the usual places.

>
> Thanks!

Never mind.


Michel Evrard

+++++++++++++++++++++++++++

>From [e-mail] (Robert Moniot)
Date: 21 Oct 1996 19:57:22 GMT
Organization: Fordham University/FCLC

The Macintosh version of ftnchek 2.9.4 built by Michel Evrard is
available:
    ftp://ftp.dsm.fordham.edu/pub/ftnchek2.9/ftnchek294.mpw.hqx

I'll leave it to others to upload this to the usual Mac sites, since I
am not well acquainted with the Mac world.

The PostScript manual is in the same ftp directory, named
ftnchek-2.9.4.ps (uncompressed) and ftnchek-2.9.4.ps.gz (gzipped).

Reports of problems with this version should probably be sent to me
rather than Evrard.
--

Bob Moniot                          email:  [e-mail]
Fordham University                        [e-mail]
Fordham College at Lincoln Center   URL:    http://www.dsm.fordham.edu/~moniot/
New York, NY  10023                 phone:  [phone]

+++++++++++++++++++++++++++

>From [e-mail] (Robert Moniot)
Date: 23 Oct 1996 01:40:47 GMT
Organization: Fordham University/FCLC

The copy of the PostScript manual for ftnchek 2.9.4 that I placed on
our ftp server on Monday was corrupted.  The copy that is there now is
correct. Sorry for any inconvenience caused by this mistake.  The URL
for it is
       ftp://ftp.dsm.fordham.edu/pub/ftnchek2.9/ftnchek-2.9.4.ps

The manual included with the source code in ftnchek-2.9.4.tar.gz was
and is OK.
--

Bob Moniot                          email:  [e-mail]
Fordham University                        [e-mail]
Fordham College at Lincoln Center   URL:    http://www.dsm.fordham.edu/~moniot/
New York, NY  10023                 phone:  [phone]

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

>From [e-mail] (WINTER  PAUL RUDOLPH)
Subject: GWorld problem
Date: 22 Oct 96 21:55:11 GMT
Organization: University of Colorado at Boulder

I noticed the tail of a thread on this topic, so I apologize for bringing
it up again, but...

I am having a problem with my GWorld. The program opens a window, creates
a GWorld, draws in the GWorld, and then opies the contents to the window.
The sequence of calls looks like this:

PlotWndw = GetNewWindow(WndwID,nil,-1)
call SetPort(PlotWndw)
myErr = NewGWorld(%ref(myFrame),0,wndwRect,nil,nil,pixPurge)
call GetGWorld(%ref(oldGWorld),%ref(oldDev))
call SetGWorld(myFrame,nil)
call EraseRect(myFrame.p^.portRect)

Plot the graph in the off-scren window

call SetGWorld(oldGWorld,oldDev)
call CopyBits(%ref(myFrame.p^.portPixMap.h^.p^),PlotWndw.p^.portBits,myFrame,p^.portRect,wndwRect,srcCopy,nil)
call SetPort(PlotWndw)

I don't dispose of the off-screen world because I use it for updating the
graph later on. I enter the above code listing at the call to GetGWorld, do
the updating (rescale the axis, highlight a peak, other things) where the
graph is plotted, and then do the SetGWorld, CopyBits, etc.

The record myFrame is kept in memory after the GWorld is created until
the window is disposed.

I also access the off-screen world to update the window in response to an
update event:

call GetPort(%ref(oldPort))
WndwPtr.p = theEvent.message
if (WndwPtr.p.eq.PlotWndw.p) then
  call SetPort(wndwPtr)
  call BeginUpdate(WndwPtr)
  call CopyBits(%ref(myFrame.p^.portPixMap.h^.p^),wndwPtr.p^.portBits,myFrame.p^.portRect,wndwRect,srcCopy,nil)
  call DrawControls(WndwPtr)
  call EndUpdate(WndwPtr)
  call SetPort(oldPort)
end if

I do this same sequence for a different window in this program, and there is
never any problem. However, with my plot window, one of the options is to
save the data. I open the standard dialog window for entering a file, and
after I close it, the region of the plot window which was behind the dialog
box doesn't draw correctly. Instead, it comes out a spolthy purple-green-red
mix. But the program doesn't crash. However, if I try to use any controls
which force a redrawing of the window, the program freezes. The sequence of
events is this:

1) Create PlotWndw and GWorld as above.
2) Save the data, which sends the code through the Update loop.
3) Try any action which forces a redraw of the window. The code executes:
   call GetGWorld(%ref(oldGWorld),%ref(oldDev))
   call SetGWorld(myFrame,nil)
   call EraseRect(myFrame.p^.portRect)
The code crashes on this call. I have checked the pointers to all the windows
etc;, they all seem to be fine. I have run the code so that it opens a
different dialog in front of the window; it runs fine on the redraw, although
it doesn't go through the Update loop first. I have opened th other window
(which goes through the same steps as this window), and it runs fine. And
the strangest part: when I do a dump of all variables (my compilier has a
call which does this), the problem goes away and the program doesn't crash.

Any thoughts? I'm working on stripping out all but the key calls so I can
try to isolate the key variables, but this could take time. Any help would
be greatly appreciated.

Thanks
Paul Winter


+++++++++++++++++++++++++++

>From [e-mail] (Jim Merkel)
Date: 22 Oct 1996 22:09:06 -0400
Organization: America Online, Inc. (1-800-827-6364)

[e-mail] (WINTER  PAUL RUDOLPH)
wrote:

[snip]
>The sequence of calls looks like this:

>PlotWndw = GetNewWindow(WndwID,nil,-1)
>call SetPort(PlotWndw)
>myErr = >NewGWorld(%ref(myFrame),0,wndwRect,nil,nil,pixPurge)
>call GetGWorld(%ref(oldGWorld),%ref(oldDev))
>call SetGWorld(myFrame,nil)
>call EraseRect(myFrame.p^.portRect)

>Plot the graph in the off-scren window

>call SetGWorld(oldGWorld,oldDev)
>call
>CopyBits(%ref(myFrame.p^.portPixMap.h^.p^),PlotWndw.p^.p>ortBits,myFrame,
p^.portRect,wndwRect,srcCopy,nil)
>call SetPort(PlotWndw)

If you are using GWorlds, you should not use GetPort() or SetPort() in
your code. Use GetGWorld() and SetGWorld() for both the GWorld and the
Window. Using SetPort() could be causing some of your problems

Jim Merkel

+++++++++++++++++++++++++++

>From [e-mail] (Indigo Moon Productions, Inc.)
Date: Thu, 24 Oct 1996 12:18:47 GMT
Organization: IgLou Internet Services (1-800-436-4456)

[e-mail] (WINTER  PAUL RUDOLPH) writes:

>I noticed the tail of a thread on this topic, so I apologize for bringing
>it up again, but...