Celestin Apprentice 5/Apprentice-Release5.iso/Information/CSMP Digest/volume 3/csmp-digest-v3-127

Text File | 1995-12-31 | 82KB | 2,102 lines

C.S.M.P. Digest             Mon, 11 Dec 95       Volume 3 : Issue 127

Today's Topics:

        Audio CD related toolbox routines?
        Changing default directory of StandardGetFile
        Creating AppleEvents
        Curious question about process mgr-A4
        Determining if a volume is local or not
        Developer VISE Code Resources?
        Experience with Zinc-Galaxy-Neuron Data?
        Floating menus sink my dialogs
        GetAppFiles - link error?
        How to get a folder's volume information!
        Random number seed dreaming.
        SAT 2.3.5 is out!
        Where to store the sprite-graphics?
        [Q] Color-Picker Popup Menu?
        [Q] How to hit-test a PICT?
        multi-player keyboard control



The Comp.Sys.Mac.Programmer Digest is moderated by Francois Pottier
([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
nef.ens.fr).  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 digest is officially distributed by two means, by email and ftp.

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-digest Your Name     Adds you to the mailing list
    signoff csmp-digest                 Removes you from the list
Once you have subscribed, you will automatically receive each new
issue as it is created.

The official ftp info is //ftp.dartmouth.edu/pub/csmp-digest.
Questions related to the ftp site should be directed to
[e-mail].

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

>From Greg Maddigan <[e-mail]>
Subject: Audio CD related toolbox routines?
Date: 22 Nov 1995 02:53:15 GMT
Organization: University of Otago, Dunedin, NZ

What toolbox routines do the apps like Audio CD Player use to obtain the
lengths of each track of an audio CD?
Are there any snippets of sample code anywhere?


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

>From [e-mail] (René Quesnel)
Date: Wed, 22 Nov 1995 17:09:04 -0500
Organization: Faculty of Music, McGill University

In article <48u3ar$[e-mail]>, Greg Maddigan
<[e-mail]> wrote:

> What toolbox routines do the apps like Audio CD Player use to obtain the
> lengths of each track of an audio CD?
> Are there any snippets of sample code anywhere?

You need:
1. The technical note about the CD-ROM driver calls at
http://www.info.apple.com/dev/technotes/Devices/dv_22.html

2. Code sample at:
http://dev.info.apple.com/source/code/Snippets/Devices/Audio_CD__/ReadMe.html

You might also want the information about the CD remote database format at:
http://www.info.apple.com/dev/technotes/Devices/dv_25.html

- ----------------------------------------------------------------------------
Rene Quesnel                                | voice:  [phone], x0523
Music Technology and Sound Recording Areas  | fax:    [phone]
Faculty of Music, McGill University         | e-mail: [e-mail]
555 Sherbrooke St.W. Montreal (QC), Canada  |
H3A 1E3                                     |
- ----------------------------------------------------------------------------

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

>From [e-mail] (Chris De Salvo)
Date: Mon, 27 Nov 1995 03:05:20 -0800
Organization: MacPlay

In article <48u3ar$[e-mail]>, Greg Maddigan
<[e-mail]> wrote:

>What toolbox routines do the apps like Audio CD Player use to obtain the
>lengths of each track of an audio CD?
>Are there any snippets of sample code anywhere?

You have to use the Device Manager calls and go through the CD-ROM
driver.  The driver calls are documented in the Mac tech notes for
devices.  It is Tech Note DV 22.

L8R
Chris

--
[e-mail]         |   Macintosh:  Changing the world,
Chris De Salvo              |        one person at a time!
Professional Mac Geek       |    -----------------------------
for MacPlay, Inc.           |      (I wish they'd hurry up!)

           http://www.deltanet.com/users/phixus

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

>From [e-mail] (John Stiles)
Subject: Changing default directory of StandardGetFile
Date: Fri, 24 Nov 1995 07:07:46 GMT
Organization: cello.gina.calstate.edu

 Hi! I am writing a program which opens an encoded file, then decodes the
contents into a new folder. The user then must pick a file from inside the
new folder. I would like to be able to set StandardGetFile to open in this
new folder, rather than forcing the user to open the new folder manually
every time. I think I have seen other programs do this--but how?

 Thanks for any help you can give...
 *Stiles

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

>From Dave Overton <[e-mail]>
Date: Sat, 25 Nov 1995 14:34:44 GMT
Organization: DataStream Imaging Systems, Inc.

CurDirStore
SFSaveDisk

are two low memory globals where the Standard File Package
stores its current DirID and vRefNum. You should be able to
pop these before calling SFGetFile.
Dave Overton

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

>From [e-mail] (Richard Wells )
Date: 27 Nov 1995 19:15:18 GMT
Organization: 3M Company

[e-mail] (John Stiles) wrote:
> Hi! I am writing a program which opens an encoded file, then decodes the
>contents into a new folder. The user then must pick a file from inside the
>new folder. I would like to be able to set StandardGetFile to open in this
>new folder, rather than forcing the user to open the new folder manually
>every time. I think I have seen other programs do this--but how?

Dave Overton <[e-mail]> responded:
>CurDirStore
>SFSaveDisk
>are two low memory globals where the Standard File Package
>stores its current DirID and vRefNum. You should be able to
>pop these before calling SFGetFile.

To do this reliably under System 7.5, you must use a standard file
hook function and modify the StandardFileReply record's sfFile
field to refer to the volume and folder you want in response to
the sfHookFirstCall message.  This is necessary under 7.5 because
users can use the General Preferences control panel to modify the
behavior of the standard file package.

_Inside Macintosh: Files_ has an example of doing this.

Any time I do this, I also save the current CurDirStore and
SFSaveDisk in the sfHookFirstCall message and restore them
during the sfHookLastCall message.

One more thing: don't forget to return sfHookChangeSelection
from the sfHookFirstCall message if you change the reply
record's sfFile field.

Yet one more thing: When you modify sfFile.parID and sfFile.vRefNum,
also set sfFile.name to an empty string.  If you leave random
garbage in there, you can crash the system.



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

>From [e-mail] (Kelly Schwarzhoff)
Subject: Creating AppleEvents
Date: Fri, 24 Nov 1995 23:28:39 -0800
Organization: Stanford University

I'm trying for the first time to send Apple Events, and can't seem to get
them to work (if it matters, I'm using Codewarrior 7).  I have the
following code:

   OSErr       err1;
   OSErr       err2;
   OSErr       err3;
   OSErr       err4;
   FInfo       fileinfo;
   AEDesc      dest;
   AppleEvent  OpenAppEvent;
   AppleEvent  Reply;


   err1 = ::HGetFInfo(mPreviewer.vRefNum, mPreviewer.parID,
mPreviewer.name, &fileinfo);
   err2 = ::AECreateDesc(typeApplSignature, &fileinfo.fdCreator,
sizeof(fileinfo.fdCreator), &dest);
   err3 = ::AECreateAppleEvent(kCoreEventClass, kAEOpenApplication, &dest,
kAutoGenerateReturnID,
    kAnyTransactionID, &OpenAppEvent);
   err4 = ::AESend(&OpenAppEvent, &Reply, kAENoReply, kAENormalPriority,
kAEDefaultTimeout, NULL, NULL);


mPreviewer is an FSSpec which holds the application's information.

The problem is that err4 returns -609, which means I'm passing an invalid
Signature.  However, I've checked it (and ran it with multiple different
programs/signatures), and the signature appears to be correct.  All of the
other err's return 0.  Suggestions?  Is there a better way to do this?  Is
this the correct way? I've never sent an apple event (and receiving them
in Codewarrior is so trivial you really never deal with the toolbox), so
I'm not sure what the correct approach is.

--
Kelly Schwarzhoff                     [e-mail]
MIME Mail is welcome

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

>From [e-mail] (Brian Stern)
Date: Sat, 25 Nov 1995 17:15:12 -0600
Organization: The University of Texas at Austin, Austin, Texas

Kelly,

You can't send an appleevent to a file; you have to send it to a running
process.  Contrary to what you were thinking you can't start an
application by sending an oapp event.  There is never any reason to send
an oapp event.  They're sent by the process manager when an app is
started.

Your code looks basically fine.  Try sending a quit event to a running
process and try using a psn to generate the address desc.

Good luck,

____________________________________________________________________
Brian  Stern  {8-{)}                                 BeBox Hairy Guy                  [e-mail]                          [e-mail]
INIT Writing FAQ etc. at <ftp://ftp.pbcomputing.com//Guests/BrianS/>

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

>From [e-mail] (Ed Anson)
Date: Sat, 25 Nov 1995 20:55:15 -0500
Organization: Tulip Software

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

> You can't send an appleevent to a file; you have to send it to a running
> process.  Contrary to what you were thinking you can't start an
> application by sending an oapp event.  There is never any reason to send
> an oapp event.  They're sent by the process manager when an app is
> started.

That is basically correct. However, it is the Finder that sends oapp when
you open an application. If you use the Process Manager to start an app,
you are obligated to send the oapp. Some applications behave rather
strangely if they don't receive that event after launch.

Incidentally, you can also use AppleScript to launch an application
without sending oapp. The Launch command does this. OTOH, simply
mentioning the application in a Tell command launches the app (if
necessary) and automatically sends oapp.

- --------------------
Ed Anson
Tulip Software
Andover, MA 01810   Check out my WWW page:
U.S.A.              <http://www.tiac.net/users/tulip/home.html>

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

>From [e-mail] (John W. Baxter)
Date: Sun, 26 Nov 1995 00:27:35 -0800
Organization: Townsend Communications

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

> There is never any reason to send
> an oapp event.  They're sent by the process manager when an app is
> started.

Actually, there is *one* reason I can think of:  to validate the behavior
of one's app when a subsequent 'oapp' comes in.  Apple seems to be
changing its mind on this...it may be, sometime in the future, that Finder
will trigger a new 'oapp' if a running app is double-clicked.  [User
testing is showing that simply bringing a left-over, windowless app to the
front may not be the best thing for the system to do.]  So the "right"
behavior may become to do the same thing whenever an 'oapp' shows up that
the app does if 'oapp' preceeds 'odoc' or 'pdoc'.  Or it may not.

Then there was an early version of FileMaker, which *attempted* to remove
its 'aevt'/'oapp' handler once one of the three events had arrived.  But
it made the call incorrectly, and the handler remained.  ["Who knows what
evil lurks in the hearts of applications?  AETracker knows..."]

   --John

--
  "This item is not available because it cannot be removed."
John W. Baxter    Port Ludlow, WA     [e-mail]

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

>From [e-mail] (Brian Stern)
Date: Sun, 26 Nov 1995 14:36:14 -0600
Organization: The University of Texas at Austin, Austin, Texas

In article <[e-mail]>, [e-mail] (Ed
Anson) wrote:

<In article <[e-mail]>,
<[e-mail] (Brian Stern) wrote:
<
<> You can't send an appleevent to a file; you have to send it to a running
<> process.  Contrary to what you were thinking you can't start an
<> application by sending an oapp event.  There is never any reason to send
<> an oapp event.  They're sent by the process manager when an app is
<> started.
<
<That is basically correct. However, it is the Finder that sends oapp when
<you open an application. If you use the Process Manager to start an app,
<you are obligated to send the oapp. Some applications behave rather
<strangely if they don't receive that event after launch.

This is not correct.  I quote from IM VI p 29-14:

"If you set this field to NIL [the launchAppParameters field], the
LaunchApplication function automatically creates and sends an Open
Application event to the launched application."

You can launch an app using the process manager call LaunchApplication.
If you do so you can specify an odoc event or any other high level event
to be the first event sent to the application.  If you don't specify any
event then the process manager sends an oapp for you.  There is never any
reason for normal applications to explicitely send an oapp event.

____________________________________________________________________
Brian  Stern  {8-{)}                                 BeBox Hairy Guy                  [e-mail]                          [e-mail]
INIT Writing FAQ etc. at <ftp://ftp.pbcomputing.com//Guests/BrianS/>

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

>From [e-mail] (Intellilink Corp.)
Subject: Curious question about process mgr-A4
Date: Tue, 14 Nov 1995 16:19:38 GMT
Organization: MV Communications, Inc.

Just wondered if anyone in netland has successfully inserted either a filter
proc or an AEEventHandler from a Code Resource ( which uses A4 ) instead
of from an application.  If you have, what gotcha's have you encountered.

I finally did manage to work with Ole on the Macintosh, in a limited and
strange and wonderful way.  Had I any choice in the matter, I would NOT
have chosen to deal with Ole at all.  Ah well.

Thanks for any info on AEHandlers living in code resources.

B. Haglind




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

>From [e-mail] (Scott Kevill)
Date: Thu, 16 Nov 1995 15:05:30 +0800
Organization: The University of Western Australia

In article <[e-mail]>, [e-mail] (Intellilink Corp.) wrote:

: Just wondered if anyone in netland has successfully inserted either a filter
: proc or an AEEventHandler from a Code Resource ( which uses A4 ) instead
: of from an application.  If you have, what gotcha's have you encountered.
:
: I finally did manage to work with Ole on the Macintosh, in a limited and
: strange and wonderful way.  Had I any choice in the matter, I would NOT
: have chosen to deal with Ole at all.  Ah well.
:
: Thanks for any info on AEHandlers living in code resources.
:
: B. Haglind

I believe it is not possible to have a code resource receive AppleEvents
since they can only be addressed to an application (or background only
application). It may be possible for a code resource to install
AEEventHandlers and receive AppleEvents meant for an application, but I
have not tried this. However sending  AppleEvents from a code resource is
fine and fairly straight-forward. What kind of AppleEvents are you wanting
to receive in a code resource?

Scott Kevill.
[e-mail]

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

>From [e-mail] (Steve Lay)
Date: Wed, 22 Nov 1995 09:41:32 +0000
Organization: DAMTP, Cambridge University

I once installed an AppleEvent handler from an XCMD running with HyperCard
and it worked fine - as long as the parent application doesn't do anything
stupid, is HLE aware and uses the AEProcessAppleEvent call you should be
alright.

However, you should check out the rules which the parent application
applies to your code to ensure that the event handler you install doesn't
move.  Eg, HyperCard may well move the code for an XCMD around in the heap
without telling you - for this reason I achieved it by loading a separate
code resource, detaching it, locking it and then installing it.  I think I
converted the handle to hex and put it in an HC global or something so
that I could clean it up when I quit.

--
Steve Lay
[e-mail]
http://www.amtp.cam.ac.uk/icrd/

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

>From [e-mail] (BradS79436)
Subject: Determining if a volume is local or not
Date: 22 Nov 1995 19:56:37 -0500
Organization: America Online, Inc. (1-800-827-6364)

How does one determine if a volume is local or remote (networked,
AppleShare, etc.) ?  PBHGetVInfo() gives me access to a bunch of
information, but that's not among it.  Do I have to come in from the
AppleShare side ?

Thanks for any help

Brad Smith
BookWorm

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

>From [e-mail] (Jump Long)
Date: 23 Nov 1995 06:05:44 -0500
Organization: America Online, Inc. (1-800-827-6364)

Brad Smith ([e-mail]) wrote:
>How does one determine if a volume is local or remote
>(networked, AppleShare, etc.) ?  PBHGetVInfo() gives me access
>to a bunch of information, but that's not among it.  Do I have
>to come in from the AppleShare side ?

You want to use PBHGetVolParms and check the vMServerAdr field of the
GetVolParmsInfoBuffer that's returned. If its zero, the volume is local.
Using the high-level calls and macros for checking GetVolParmsInfoBuffer
fields in the MoreFiles sample code makes this rather easy as shown here:

Boolean IsNetworkVolume(StringPtr volName, short vRefNum)
{
 GetVolParmsInfoBuffer volParmsInfo;
 long infoSize;

 infoSize = sizeof(GetVolParmsInfoBuffer);
 if ( HGetVolParms(volName, vRefNum, &volParmsInfo, &infoSize) == noErr )
 {
  /* If volume supports GetVolParms, see if it is a network volume */
  return ( isNetworkVolume(volParmsInfo) );
 }
 else
 {
  /* Volume doesn't support GetVolParms, so it isn't a network volume */
  return ( false );
 }
}

HGetVolParms is a MoreFiles function and isNetworkVolume is a MoreFile
macro.

- Jim Luther

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

>From [e-mail] (Christine van Veen)
Subject: Developer VISE Code Resources?
Date: Sun, 26 Nov 1995 16:19:01 -0400
Organization: Bell-Northern Research/Northern Telecom

I have good knowledge of C programming in the PC and Unix environment.  I
have been attempting to learn to write code resources in C for use with
Developer VISE - with very little success.  Can anyone explain to me what
is necessary to write a code resource for Developer VISE?

Currently, I am working on writing a code resource to "age" a file.  This
means that, before a file is installed, the code resource will:

* search for the file to be installed, in case it already exists
* if the file exists, rename the file to some unique name
      e.g.  "TheFile"  might become  "TheFile.11.32"
               if the file is renamed at 11:32 a.m.

If anyone has experience with this, I would really appreciate hearing any
suggestions.

- Christine van Veen

**********************************************
  Christine van Veen       [e-mail]
  Mac/Unix Systems Integration
  Bell Northern Research
  Phone:  [phone]   Fax: [phone]
**********************************************

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

>From [e-mail] (David Walton)
Date: Mon, 27 Nov 1995 18:46:47 -0500
Organization: University of Notre Dame

In article <cvanveen-2611951619010001@47.199.32.68>, [e-mail]
(Christine van Veen) wrote:

>I have good knowledge of C programming in the PC and Unix environment.  I
>have been attempting to learn to write code resources in C for use with
>Developer VISE - with very little success.  Can anyone explain to me what
>is necessary to write a code resource for Developer VISE?
>
>Currently, I am working on writing a code resource to "age" a file.  This
>means that, before a file is installed, the code resource will:
>
>* search for the file to be installed, in case it already exists
>* if the file exists, rename the file to some unique name
>      e.g.  "TheFile"  might become  "TheFile.11.32"
>               if the file is renamed at 11:32 a.m.
>
>If anyone has experience with this, I would really appreciate hearing any
>suggestions.

Here's a very brief external that I use to check for the existence of
OpenTransport in an installer I wrote.  It's pretty straightforward:  just
link with the right libraries, get the right parameters, and you're home
free.  I have written something that vaguely resembles what you're
specifically looking for, so if you'd like to discuss more details, e-mail
me.

(Also, please excuse the ugliness of the code.)

David Walton



#ifndef __EXTERNCODEDEFINES__
#include <ExternCodeDefines.h>
#endif
#include <Gestalt.h>
#ifndef gestaltOpenTransportAttr
#define gestaltOpenTransportAttr 'otan'
#endif

#define kTestHasOT      1
#define kTestHasMacTCP  2

pascal void main (ExternParmBlock *PB);



/* ====================================================================

   main

   ==================================================================== */

pascal void main (ExternParmBlock *PB)
{
   long     gestaltValue;
   Boolean  hasOT = false;
   OSErr    error;

   error = Gestalt(gestaltOpenTransportAttr, &gestaltValue);
   if (!error && gestaltValue != 0)
      hasOT = true;
   switch (PB->refCon)
   {
      case kTestHasOT:
         PB->installFork = (hasOT  ?  PB->installFork  :  kDontInstallFile);
         break;
      case kTestHasMacTCP:
         PB->installFork = (hasOT  ?  kDontInstallFile  :  PB->installFork);
         break;
      default: /* This should never be called  */
         PB->installFork = kDontInstallFile;
         break;
   }
}

--
David Walton                    |  Macintosh Consultant & Graduate Student
Mailto:[e-mail]   |  Office of Information Technologies
http://www.nd.edu/~dwalton1/    |  Dept. History & Philosophy of Science
Voice: 219-631-8024             |  University of Notre Dame

   "Beware the advice of successful people; they do not seek company."

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

>From [e-mail] (Christopher R. Bailey)
Subject: Experience with Zinc-Galaxy-Neuron Data?
Date: Tue, 14 Nov 1995 09:45:02 -0800
Organization: Beetle's Sprawl

I am conducting an investigation into cross platform frameworks for my
company.  We have an existing product that is cross platform (Mac &
Windows) using our own cross platform framework.  We'd like to move to a
commercial product to attain more GUI features, not have to worry about
maintaining the framework ourselves, possibly pick up international
support and database connectivity, etc.

I started with a list including Galaxy, Neuron Data, XVT, Zinc, StarView,
ParcPlace VisualWorks, zApp, C++/Views, Utah, MFC, ODF, and Prograph.  We
require that Mac, PowerMac, Windows 95, Windows NT (Intel and preferably
Alpha), and possibly Windows 3.1 be supported, and that the product be a
C++ product, preferably supporting a MVC paradigm.  Our initial
investigation included two non-C++ products as they seemed to have enough
compelling arguments to look at them.

As of now, we've narrowed the field down to Neuron Data's Open Interface
Elements (and possibly C/S Elements), Visix's Galaxy, and Zinc's Zinc
Application Framework.  With Neuron Data, we will base our evaluation on
the next release (March/April '96) as it's claimed to be a much better C++
product.  We are also interested in Visix's upcoming Galaxy Global
(version for creating international/Unicode/multi-byte apps), as well as
Zinc's international offering.  At this time price is not a major factor
in the decision.  Also we probably won't be making a decision for several
months, which allows us to look at the upcoming releases from some of
these companies.

What I'd like to find out is who has chosen one of these products, and for
what reasons.  More importantly, I'd like to find people who've chosen
Zinc/Galaxy/Open Interface when comparing these three (and/or others).

Any information you feel like providing would be much appreciated.  Please
email me as well ([e-mail]).  Specifically I'd be
interested in your opinions of the companies, the tech support, the
features of the product, use of the GUI design tool, experience using I18N
features/abilities (which I guess mainly pertains to Zinc at this point),
documentation, training, etc.  Thanks in advance!

_____________ Christopher R. Bailey _____________
[e-mail]
http://www.quake.net/~baileyc
Macintosh, for those who can see through Windows.
Ride fast, take chances!

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

>From [e-mail] (Sean A Corfield)
Date: Wed, 15 Nov 1995 12:11:57 +0000
Organization: OCS

In article <[e-mail]>,
[e-mail] (Christopher R. Bailey) wrote:

|> What I'd like to find out is who has chosen one of these products, and
for
|> what reasons.  More importantly, I'd like to find people who've chosen
|> Zinc/Galaxy/Open Interface when comparing these three (and/or others).

I worked for a company that went through a similar evaluation process and
picked Zinc. Neuron Data was simply too expensive -- it had a run-time
licence at the time. Galaxy was too all-encompassing for what we needed
(just a GUI, no O/S service layer). Zinc has some peculiar C++ idioms and a
steep learning curve (not as steep as Galaxy, I'd say). However, there were
many, many problems with the Motif version of Zinc and I just heard that my
old company has junked the project (and reverted to custom-building their
Motif and Windows interfaces). Zinc admitted that their Motif port was very
buggy and claimed that the forthcoming v5.x will be enough of a rewrite to
solve this.

I have since heard of similar problems with Zinc in a Windows environment
(my old company never got as far as completing the Windows port to discover
whether this is true or not).

Sean A. Corfield
Technical Director
Object Consultancy Services




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

>From [e-mail] (Gerbert Orasche)
Date: 17 Nov 1995 21:53:47 GMT
Organization: IICM

In article <[e-mail]>, [e-mail]
says...
>
>I am conducting an investigation into cross platform frameworks for my
>company.  We have an existing product that is cross platform (Mac &
>Windows) using our own cross platform framework.  We'd like to move to a
>commercial product to attain more GUI features, not have to worry about
>maintaining the framework ourselves, possibly pick up international
>support and database connectivity, etc.
I am cross-developing for Windows 3.x,95,NT,Mac, and PowerMac using
Microsofts tools. I am now porting an VRML viewer to Mac, with the cross
development toolkit for Macintosh. It works quite fine (full MFC support),
but I do not now very much about performance yet.

cu
--
Gerbert Orasche at Graz University Of Technology
IICM (Institute for Information Processing and Computer Supported New
Media)
Inter-Net [e-mail]
          [e-mail]
Tel-Net   ++43-316-873-5607
http://hyperg.iicm.tu-graz.ac.at/TU-5060.Pers.gorasche


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

>From [e-mail] (Brett Kuehner)
Date: Sat, 18 Nov 1995 04:24:55 GMT
Organization: Internet Online Services

On Tue, 14 Nov 1995 09:45:02 -0800, [e-mail] (Christopher R.
Bailey) wrote:

>I am conducting an investigation into cross platform frameworks for my
>company.

I did very much the same thing, around a year and a half ago. I looked
at Zinc, Visix Galaxy, Neuron Data, zApp, and a half-dozen others
(most of the ones on your list). The main contenders I saw were Galaxy
and Neuron Data, mostly because they seemed to be well-designed, and
not just a pile of C++ classes. They both also had some kind of
scripting support. However, they were both very expensive (thousands
of dollars per developer seat).

I liked Galaxy quite a bit, and a demo that I saw of their interface
builder leads me to believe their builder is (was a year ago, anyway)
far superior to Neuron Data, supporting a nice visual layout mechanism
of springs and struts, among other things. For their other stuff,
Galaxy and Neuron Data both seemed to have their good and bad points.
Neither had mature C++ support at the time, although this seems to
have improved for both.

My final decision was to get Neuron Data. This was based mainly on two
factors:
1) Neuron Data was cheaper than Galaxy (by a thousand or two per seat,
at least)

2) Most importantly, Neuron Data offered an evaluation copy that made
it easy to see it it was suitable for my project. They were very easy
to deal with (my salesperson, Jenny Leipziger, was very helpful. I
recommend her if you want any questions answered), and they gave me
eval copies for both Sun Sparc and PC/Windows without any hassle.

Visix, on the other hand, would not give me a demo no matter how hard
I tried. The best I could do was to get a set of manuals (C manuals,
at that. They said the C++ manuals weren't ready yet). To check out
the Galaxy features, Visix recommended taking a week-long course at
their site in Virginia. This is hardly practical, not even considering
the expense of travel and lodging, since trying to get away from work
for a week just for an evaluation is kind of extreme (and impossible).

Anyhow, Neuron Data worked out pretty well for the project that I was
leading (real-time financial information display on 486/66 PC's
running Windows 3.11). Tech support was pretty responsive, although
they took a while to respond to some of the bugs. Graphics performance
was good, once we figured out the way to optimize drawing.
The event model ND uses is good, and made it easy to have an event
architecture that would have been difficult to do in plain windows. It
also did not seem to interfere with using Windows API calls intermixed
in the code, so you can get at DDE and other functions which are (or
were) not supported directly by ND.
Neuron Data did not support some of the compilers that I wanted to use
uner Windows, so check to see that support for your preferred compiler
is available.
The app also seemed to port to the Sparc OK, although I didn't see the
completion of that effort.

Overall, I was happy with the decision to use Neuron Data, although
for certain types of projects I might choose something else. For
example, I'm now doing Windows and NT development that does not have
to be portable, and Borland's Delphi is a great environment for this.


I left the company where I was using ND, so I'm not sure what the
current state is. However, I have a friend at another company who was
using Neuron Data, and abandoned it, as he needed support for a
Windows 95 look-and-feel, and ND wasn't planning to provide one in the
near-term. If this is important to you, I recommend quizzing them
about delivery schedules (and taking them with a grain of salt, as
they were late on the C++ support)

Well, if you made it through all that, I also have a brief comment on
Zinc and zApp. I thought they were both poorly structured (in C++
class hierarchy terms), and that their interface builders were also
terrible. They have probably improved some by now, though.

If you have any questions, let me know.

                Brett


--
Brett Kuehner
[e-mail]


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

>From [e-mail] (Anthony D. Tribelli)
Date: Mon, 20 Nov 1995 18:25:59 GMT
Organization: NETCOM On-line Communication Services (408 261-4700 guest)

Gerbert Orasche ([e-mail]) wrote:
: I am cross-developing for Windows 3.x,95,NT,Mac, and PowerMac using
: Microsofts tools...
: ... It works quite fine (full MFC support),
: but I do not now very much about performance yet.

I recently used MSVC++ for Mac 2.0 (I haven't received 4.0 yet) to create
a simple 3D molecular visualization program. As you say, MFC was
supported very well. The rendering performance of my program was poor
compared to our native Mac applications (built with Metrowerks
CodeWarrior). By using conditional compilation I was able to get my
program to run 2.5 TIMES FASTER. It now runs at a speed equivalent to our
native Mac applications.

The changes were relatively trivial. I replaced two Windows GDI calls with
Apple QuickDraw calls: Ellipse and Polygon => PaintOval and PaintPoly. I
also removed some unneccessary Windows GDI calls from Mac builds:
CreatePen, CreateBrush, SelectObject.

Tony

--
- ----------------
Tony Tribelli
[e-mail]

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

>From [e-mail] (John Downing)
Date: Mon, 27 Nov 1995 09:34:09
Organization: NASA Goddard Space Flight Center -- Greenbelt, Maryland USA

In article <[e-mail]> [e-mail] (Anthony D. Tribelli) writes:
>From: [e-mail] (Anthony D. Tribelli)
>Subject: Re: Experience with Zinc/Galaxy/Neuron Data?
>Date: Mon, 20 Nov 1995 18:25:59 GMT

>Gerbert Orasche ([e-mail]) wrote:
>: I am cross-developing for Windows 3.x,95,NT,Mac, and PowerMac using
>: Microsofts tools...
>: ... It works quite fine (full MFC support),
>: but I do not now very much about performance yet.

>I recently used MSVC++ for Mac 2.0 (I haven't received 4.0 yet) to create
>a simple 3D molecular visualization program. As you say, MFC was
>supported very well. The rendering performance of my program was poor
>compared to our native Mac applications (built with Metrowerks
>CodeWarrior). By using conditional compilation I was able to get my
>program to run 2.5 TIMES FASTER. It now runs at a speed equivalent to our
>native Mac applications.

>The changes were relatively trivial. I replaced two Windows GDI calls with
>Apple QuickDraw calls: Ellipse and Polygon => PaintOval and PaintPoly. I
>also removed some unneccessary Windows GDI calls from Mac builds:
>CreatePen, CreateBrush, SelectObject.

Doesn't this sort of defeat the purpose?

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

>From [e-mail] (Brian Quinlan)
Subject: Floating menus sink my dialogs
Date: 22 Nov 1995 20:09:35 GMT
Organization: Simon Fraser University

I just finished writting some code to handling floating windows. It works
my using the low-level window manager routines. My problem is that
the dialog manager seems to check to see if it is the front window during
DialogSelect and IsDialogEvent.  It occured to me that I could use two
strategies to overcome this problem. First I could patch FrontWindow( ) and
hope that that's what those routines use. Second I could fudge the window
list just before calling those routines to make the dialog appear to be
the front window. What should I do? Is there a third and more elegant
technique?

--

Brian Quinlan
[e-mail]

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

>From [e-mail] (David Sinclair)
Date: 23 Nov 1995 20:16:10 GMT
Organization: iconz, Auckland, New Zealand

Brian Quinlan ([e-mail]) wrote:
: I just finished writting some code to handling floating windows. It works
: my using the low-level window manager routines. My problem is that
: the dialog manager seems to check to see if it is the front window during
: DialogSelect and IsDialogEvent.  It occured to me that I could use two
: strategies to overcome this problem. First I could patch FrontWindow( ) and
: hope that that's what those routines use. Second I could fudge the window
: list just before calling those routines to make the dialog appear to be
: the front window. What should I do? Is there a third and more elegant
: technique?

I was struggling with this very same issue just earlier this week,
including discussing it with the author of the Infinity Windoid and others.

My solution was to use the GhostWindow low-mem global: it prevents
FrontWindow from returning a floating window as frontmost.  It works, but
the problem is that it only allows you to have one floating window.
That's fine for me, but if you need more, I'd suggest fiddling with the
window list (or rather, replace the WindowList low-mem global with the
window you want to consider frontmost, then restore the old value
afterwards.  I haven't tried using this method, but it should work.

Personally, I consider manipulating documented low-mem globals to be
better than patching routines -- less likely to cause compatibility
problems.

Good luck!  (And if you do find some better technique, please let me know.)

David Sinclair
Dejal Userware

--

|\  ._   .  _  .     Dejal Userware  |  [e-mail]  (preferred)
| | |_   | |_| |     PO Box 33-1011  |  [e-mail], [e-mail]
|/  |_ \_| | | |_    Takapuna        |  [e-mail] (avail shortly)
- ---------------    Auckland 1309   |  [e-mail]
 U S E R W A R E     New Zealand     |  ftp://ftp.amug.org/pub/contrib/dejal/
- ---------------                    |  http://pobox.com/~dejal/

Author of SndConverter Pro, SndPlayer, QuickEncrypt, TextMerge, TextSplitter
and other Shareware and Freeware products for the Macintosh.

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

>From [e-mail] (Cameron Esfahani)
Date: Thu, 23 Nov 1995 18:51:23 -0800
Organization: Apple Computer, Inc.

If I could make a suggestion.  Why don't you consider using Dean Yu's
Floating Window code.  It doesn't patch anything or mess with any
lowmems.  It was introduced in Develop 15 in an article with sample code.
The latest developer CD would have the latest and greatest version of the
code.  Most people that I know who do floating windows use this code.

Cameron Esfahani

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

>From [e-mail] (David Sinclair)
Date: 24 Nov 1995 06:22:49 GMT
Organization: iconz, Auckland, New Zealand

Cameron Esfahani ([e-mail]) wrote:
: If I could make a suggestion.  Why don't you consider using Dean Yu's
: Floating Window code.  It doesn't patch anything or mess with any
: lowmems.  It was introduced in Develop 15 in an article with sample code.
: The latest developer CD would have the latest and greatest version of the
: code.  Most people that I know who do floating windows use this code.

Simple -- because it doesn't support modeless dialogs in the way I (and
presumably the original poster) need.  The program I'm working on (5Mb of
source code!) uses modeless dialogs all over the place as normal windows
(since it needs lots of entry fields, well suited to the Dialog Manager).
The way the Develop 15 code works is to treat modeless dialogs as modal,
and bring them up in front of the floating window(s), which is no good
for me.

The problem is that IsDialogEvent and DialogSelect both call FrontWindow
internally, which breaks the Develop 15 code if a modeless dialog is to
be treated as a normal window.  Only by using ghostWindow or fiddling
with the window list can that be avoided, unfortunately.

Regards,

David Sinclair
Dejal Userware

--

|\  ._   .  _  .     Dejal Userware  |  [e-mail]  (preferred)
| | |_   | |_| |     PO Box 33-1011  |  [e-mail], [e-mail]
|/  |_ \_| | | |_    Takapuna        |  [e-mail] (avail shortly)
- ---------------    Auckland 1309   |  [e-mail]
 U S E R W A R E     New Zealand     |  ftp://ftp.amug.org/pub/contrib/dejal/
- ---------------                    |  http://pobox.com/~dejal/

Author of SndConverter Pro, SndPlayer, QuickEncrypt, TextMerge, TextSplitter
and other Shareware and Freeware products for the Macintosh.


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

>From [e-mail] (Martin-Gilles Lavoie)
Date: Sat, 25 Nov 1995 10:17:13 -0500
Organization: ZERCOM Technologies Inc.

In article <49001v$[e-mail]>, [e-mail] (Brian
Quinlan) wrote:

> I just finished writting some code to handling floating windows. It works
> my using the low-level window manager routines. My problem is that
> the dialog manager seems to check to see if it is the front window during
> DialogSelect and IsDialogEvent.  It occured to me that I could use two
> strategies to overcome this problem. First I could patch FrontWindow( ) and
> hope that that's what those routines use. Second I could fudge the window
> list just before calling those routines to make the dialog appear to be


> the front window. What should I do? Is there a third and more elegant
> technique?

Yes: rewrite the dialog manager.

No kidding, it's essentially what I did.  I was using the WindowsExtension.h/.c
 library as found and documented in Develop (and also available womewhere
in ftp.apple.com, I think), but found that it didn't support dialog-based
floaters, and that the Dialog Manager didn't support floaters.  I reworked
the WindowsExtensions lib to use a different "windowKind" variable
("myWindowKind", to be more exact and original..), and for that lib to use
DialogRecords instead of WindowRecords (plain windows take a little more
mem, but the tradeoff is acceptable).  Then, I wrote "DialogSelect" and
"IsDialogEvent" equivalents to replace the dialog manager.  This is
simple, and I beleive I have posted the code a while ago.  (I then whent
on to rewrite DrawDialog to support unlimited font/size/face/justification
and a bunch other cool thing, like drag-and-drop, colored background in
stat text, cursor management, etc).

HEY APPLE!  IF YOU WANT A NATIVE DIALOG MANAGER, I MOSTLY HAVE IT!!!!

Just teasing.

MGL

Martin-Gilles Lavoie
- -------------------------------------------------------------------
No!...try not.  Do, or do not.  There is no try.
    --Yoda on error handling.

Wars does not make one great.
    --Yoda on "Great Warrior"

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

>From [e-mail] (Francois Pottier)
Date: 26 Nov 1995 20:13:09 GMT
Organization: Ecole Normale Superieure, Paris

In article <mouser-2511951017130001@204.191.6.123>,
Martin-Gilles Lavoie <[e-mail]> wrote:

>HEY APPLE!  IF YOU WANT A NATIVE DIALOG MANAGER, I MOSTLY HAVE IT!!!!

They don't - it's going away with Copland.

Seriously though, you're right, bypassing the Dialog Manager is the
right way. That's the way PowerPlant does floating windows, for
instance. It never uses a single dialog.



--
Francois
[e-mail]
http://www.eleves.ens.fr:8080/home/pottier/

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

>From [e-mail] (Thomas Reed)
Date: Mon, 27 Nov 1995 11:14:29 -0600
Organization: Washington University

In article <49001v$[e-mail]>, [e-mail] (Brian
Quinlan) wrote:

>I just finished writting some code to handling floating windows. It works
>my using the low-level window manager routines. My problem is that
>the dialog manager seems to check to see if it is the front window during
>DialogSelect and IsDialogEvent.

I've heard many people say they've had this problem -- I don't know what
the cause is, but my floating windows work perfectly and don't have this
problem.  Get hold of a copy of Matt Slot's appe Windows code.  That's
what my floaters are based on, and it's very easy to reuse.

-Thomas

=====================================================
Thomas Reed                     Washington University
[e-mail]               Medical School
[e-mail]            Saint Louis, MO
http://medinfo.wustl.edu/~reed
- ---------------------------------------------------
Clothes make the man.  Naked people have little or no
influence on society.  -- Mark Twain
=====================================================

Opinions posted are not the opinions of Wash. U.

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

>From [e-mail] (Julian Harris)
Subject: GetAppFiles - link error?
Date: Thu, 23 Nov 1995 23:51:36 +0100
Organization: Computer Science, The University of Auckland

Hi all.

I'm compiling this rather old source that has references to _GetAppFiles_
and _CountAppFiles_.

I want this program to work under System 6 _and_ PowerMac so does anyone
know where these traps have disappeared to?

TIA.

--
.  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .  .
Microsoft is not the answer.                    >  Julian Harris, Programmer   >
Microsoft is the question.                     >  Comp. Sci. Dept.     x8915  >
                                              >  The University of Auckland  >
NO is the answer.                            >  [e-mail]    >

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

>From [e-mail] (Peter S. Lau)
Date: 23 Nov 1995 15:34:06 GMT
Organization: Internex Online, Toronto, Ontario, Canada (416 363 3783)

In article <[e-mail]> [e-mail] (Julian Harris) writes:

>I'm compiling this rather old source that has references to _GetAppFiles_
>and _CountAppFiles_.
>
>I want this program to work under System 6 _and_ PowerMac so does anyone
>know where these traps have disappeared to?

A quick solution is to open SegLoad.h and copy those funtion
declarations into your file where GetAppFiles() resides... if you look
at SegLoad.h, it's excluded when compiling ppc...

Not sure what will happen in the future...

pete


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

>From [e-mail] (John W. Baxter)
Date: Thu, 23 Nov 1995 13:13:48 -0800
Organization: Townsend Communications

In article <[e-mail]>,
[e-mail] (Julian Harris) wrote:

> Hi all.
>
> I'm compiling this rather old source that has references to _GetAppFiles_
> and _CountAppFiles_.
>
> I want this program to work under System 6 _and_ PowerMac so does anyone
> know where these traps have disappeared to?

It's time to redo the app to support the "required events".  Copy the
source out of Inside Mac: IAC.  Or any modern sample app.

   --John

--
  "This item is not available because it cannot be removed."
John W. Baxter    Port Ludlow, WA     [e-mail]

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

>From [e-mail] (Francois Pottier)
Date: 24 Nov 1995 14:08:34 GMT
Organization: Ecole Normale Superieure, Paris

In article <[e-mail]>,

>I want this program to work under System 6 _and_ PowerMac so does anyone
>know where these traps have disappeared to?

To the bit bucket. These calls *do not exist* on Power Macs.

--
Francois
[e-mail]
http://www.eleves.ens.fr:8080/home/pottier/

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

>From [e-mail] (Chris De Salvo)
Date: Mon, 27 Nov 1995 03:00:40 -0800
Organization: MacPlay

In article <[e-mail]>,
[e-mail] (Julian Harris) wrote:

>I'm compiling this rather old source that has references to _GetAppFiles_
>and _CountAppFiles_.
>
>I want this program to work under System 6 _and_ PowerMac so does anyone
>know where these traps have disappeared to?

That mechanism does not exist under the PowerPC version of the system
software.  You'll need to install AppleEvent handlers for the 'odoc' (open
document) event.

Do you really NEED to support System 6?  Not many copiesof that still in use.

L8R
Chris

--
[e-mail]         |   Macintosh:  Changing the world,
Chris De Salvo              |        one person at a time!
Professional Mac Geek       |    -----------------------------
for MacPlay, Inc.           |      (I wish they'd hurry up!)

           http://www.deltanet.com/users/phixus

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

>From [e-mail] (Anders Wahlin)
Subject: How to get a folder's volume information!
Date: Tue, 21 Nov 1995 14:32:06 GMT
Organization: Hum Fak:s Dataservice

In my program I let the user select a folder (using CustomGetFolder). Then
the program will get the FSSpec of that folder and then it will store it
in an 'alis'-resource. Now, if that folder is located on an AppleShare
volume, the user has to enter his/hers user name and password to have
access to it the next time.

I know how most of this is done except for:

* How can I determine a volume's zoneName and serverName from a folder's FSSpec
* How do I know that the selected folder (FSSpec) is located on an
AppleShare volume.

I'd really appreciate all kind of help!

Many thanks

--
Anders Wahlin
[e-mail]

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

>From [e-mail] (Marshall Clow)
Date: Thu, 23 Nov 1995 00:55:49 -0800
Organization: Aladdin Systems

In article <[e-mail]>,
[e-mail] (Anders Wahlin) wrote:

>In my program I let the user select a folder (using CustomGetFolder). Then
>the program will get the FSSpec of that folder and then it will store it
>in an 'alis'-resource. Now, if that folder is located on an AppleShare
>volume, the user has to enter his/hers user name and password to have
>access to it the next time.
>
>I know how most of this is done except for:
>
>* How can I determine a volume's zoneName and serverName from a folder's FSSpec
>
Call PBHGetVolMountInfo.

>* How do I know that the selected folder (FSSpec) is located on an
>AppleShare volume.
>
Telling if a drive is local or remote is easy; call PBHGetVolParms and
check to see if buffer.vMServerAdr != 0.

BTW, use the vRefNum in the FSSpec to refer to the correct volume.

--
Marshall Clow
Aladdin Systems

"They that can give up essential liberty to obtain a little temporary
 safety deserve neither liberty nor safety." -- Benjamin Franklin
 _Historical Review of Pennsylvania_, 1759

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

>From [e-mail] (Jump Long)
Date: 23 Nov 1995 06:05:09 -0500
Organization: America Online, Inc. (1-800-827-6364)

Anders Wahlin wrote:
>In my program I let the user select a folder (using
>CustomGetFolder). Then the program will get the FSSpec of that
>folder and then it will store it in an 'alis'-resource. Now, if
>that folder is located on an AppleShare volume, the user has to
>enter his/hers user name and password to have access to it the
>next time.
>
>I know how most of this is done except for:
>
>* How can I determine a volume's zoneName and serverName from a
>folder's FSSpec * How do I know that the selected folder
>(FSSpec) is located on an AppleShare volume.
>
>I'd really appreciate all kind of help!

You seem to be a little confused...

Let's say you get an FSSpec to a folder from your CustomGetFolder routine.
If you want to store that location for later use, you should create an
alias to the folder by passing the FSSpec to NewAlias. NewAlias will
create an AliasHandle that contains the alias record. The alias record
contains everything needed to find that folder later. You can then add the
AliasHandle to a resource file as an 'alis' resource.

Later, when you need to find the folder, get the 'alis' resource and pass
that resource handle to ResolveAlias. ResolveAlias (if successful) will
return an FSSpec to the folder and it needed, will take care of any user
interface needed to mount the volume the folder is on.

Going back to your question:
>* How can I determine a volume's zoneName and serverName from a
>folder's FSSpec * How do I know that the selected folder
>(FSSpec) is located on an AppleShare volume.

You can get that information the same way the Alias Manager gets it - call
PBGetVolMountInfoSize and PBGetVolMountInfo. If PBGetVolMountInfoSize and
PBGetVolMountInfo return paramErr (-50), it means that the volume doesn't
support those calls. If they return noErr (0), it might be an AppleShare
volume. You can check the media field of the VolMountInfo record returned
by PBGetVolMountInfo to get the type of volume - 'afpm' is what you'll get
for AppleShare volumes.

- Jim Luther

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

>From [e-mail] (John Maling)
Subject: Random number seed dreaming.
Date: 12 Nov 1995 19:48:55 GMT
Organization: Zip Nada and Zero

OK, this must be a weird one, it came to me in a dream. I know that the
random number generator is pseudo-random, that if you initilise it with
the same number you get the same result sequence. So, could this be used
to play back an identical sequence of pseudo-random events inside a game
reguardless of the individual Mac the game runs on? Or does the random
number generator on each Mac produce different values even if the seed is
the same?

I need to create a record of complex game events to be sent to certain
players, events which are themselves independant of player input. So if
the ramdom seed produced the same result on every Mac then I would only
need to pass the seed, and use it to reproduce the event sequence, rather
than send the entire event sequence.

From: [e-mail]

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

>From [e-mail] (Juri Munkki)
Date: 13 Nov 1995 03:28:25 GMT
Organization: Helsinki University of Technology

In article <[e-mail]> [e-mail] (John Maling) writes:
>OK, this must be a weird one, it came to me in a dream. I know that the
>random number generator is pseudo-random, that if you initilise it with
>the same number you get the same result sequence. So, could this be used
>to play back an identical sequence of pseudo-random events inside a game
>reguardless of the individual Mac the game runs on? Or does the random
>number generator on each Mac produce different values even if the seed is
>the same?

I think all Macs use the same algorithm for Random, but you shouldn't
trust in this. If you need a repeatable set of pseudorandom numbers,
you should use your own version of Random.

Books on computer algorithms usually have a chapter on pseudorandom
number generation.

>I need to create a record of complex game events to be sent to certain
>players, events which are themselves independant of player input. So if
>the ramdom seed produced the same result on every Mac then I would only
>need to pass the seed, and use it to reproduce the event sequence, rather
>than send the entire event sequence.

Arashi uses two random generators: one is used for visual effects and
stuff that doesn't affect the game engine. The other one is used for
internal stuff. It plays back saved games when it is idle. The saved
games consist of the random seed and only a few bits of information per
frame: movement and shoot+zap flags. This means that it takes only 20
bytes/second (or 72 KB/hour) to save games.

Marathon saves films this way, but I don't know how much randomness they
have in the game. In many cases random numbers are not necessary in games,
because player actions are different enough every time, so they could
actually probably write the whole game without using a single random number...

--
Juri Munkki [e-mail]      There ain't no such thing as a shareware lunch.
http://www.iki.fi/~jmunkki              Windsurfing: Faster than the wind.

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

>From [e-mail] (David Wareing)
Date: Tue, 21 Nov 1995 14:18:31 +1100
Organization: Weyland Yutani

In article <[e-mail]>,
[e-mail] (John Maling) wrote:

>OK, this must be a weird one, it came to me in a dream. I know that the
>random number generator is pseudo-random, that if you initilise it with
>the same number you get the same result sequence. So, could this be used
>to play back an identical sequence of pseudo-random events inside a game
>reguardless of the individual Mac the game runs on? Or does the random
>number generator on each Mac produce different values even if the seed is
>the same?

I'd think that it would produce the same value on any Mac, but I guess it
would depend upon whether _Random has changed between System releases
and whether or not the PPC version is different from the 68K code.

As Juri Munkki pointed out recently, it'd be safer to write your own
Random function and use that for all seeding. Do a search through the
csmp digests for dozens of articles on rolling your own Random function.

--
David Wareing                                 [e-mail]
Belair, South Australia           http://www.AmbrosiaSW.com/~dwareing/
Macintosh Games & Multimedia Programming


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

>From [e-mail] (Francois Pottier)
Date: 24 Nov 1995 13:31:47 GMT
Organization: Ecole Normale Superieure, Paris

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

>I'd think that it would produce the same value on any Mac, but I guess it
>would depend upon whether _Random has changed between System releases
>and whether or not the PPC version is different from the 68K code.

This has been discussed before, and someone from Apple said that they
were tempted to change the algorithm when rewriting Random() for
PowerPC, but they didn't because some applications relied on it. So I
guess it would be safe to assume it will never change. Look in the
digests for this thread.


--
Francois
[e-mail]
http://www.eleves.ens.fr:8080/home/pottier/

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

>From [e-mail] (Gordon Henriksen)
Subject: SAT 2.3.5 is out!
Date: 22 Nov 1995 03:14:23 GMT
Organization: Micron Internet Services

Doesn't the subject say it all?

ftp://mirrors.aol.com/pub/info-mac/rec/sprite-anim-kit-235.hqx

                      TTFN,
                        Gordon Henriksen
                        [e-mail]
                 _________________________________
                 Q: Where do you want to go today?
                     A: Anywhere but Chicago!
__________________________________________________________________
Microsoft Network is prohibited from redistributing this work in
any form, in whole or in part. Copyright, Gordon Henriksen, 1995.
License to distribute this post is available to Microsoft for $150.
Posting without permission constitutes an agreement to these terms.
Please send notices of violation to [e-mail] and
[e-mail]

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

>From [e-mail] (Ingemar Ragnemalm)
Date: 24 Nov 1995 17:01:09 GMT
Organization: (none)

[e-mail] (Gordon Henriksen) writes:

>Doesn't the subject say it all?

>ftp://mirrors.aol.com/pub/info-mac/rec/sprite-anim-kit-235.hqx

Thanks! :-) It is strange - my announcements seem not to have shown up yet.

There is a little more to say, like that this is the first version that I
have successfully used with the PPC version of SC++ (though I would recommend
some other compiler to anyone with less than 20-30 megs of RAM). Also, it
comes in two parts nowadays. The demo programs are in a separate archive.

There is a bug fix coming up too, as mentioned in another thread.
It is already available at Lysator (ftp.lysator.liu.se) and should show
up at Info-Mac and Umich very soon.

--
- -
Ingemar Ragnemalm, PhD
Image processing, Mac shareware games
E-mail address: [e-mail] or [e-mail]

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

>From [e-mail] (Thorsten Goertz)
Subject: Where to store the sprite-graphics?
Date: 20 Nov 1995 15:32:45 GMT
Organization: University of Hamburg -- Germany

I'm currently writing a program that has to display some
large picts in 256 colors and animates some smaller picts
as sprites on the screen.
The picts are stored in the resource-fork. Reading and
copying them to an offscreen-GWorld is very slow!
What is the preffered way for storing and reading the
graphics without waiting 10-15 seconds (at least on
my LC475) before the animation starts?

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

>From [e-mail] (Mark Chrisman)
Date: Mon, 20 Nov 1995 17:03:30 -0800
Organization: Inreach's InterNetNews Site

In article <48q72t$[e-mail]>,
[e-mail] (Thorsten Goertz) wrote:

|I'm currently writing a program that has to display some
|large picts in 256 colors and animates some smaller picts
|as sprites on the screen.
|The picts are stored in the resource-fork. Reading and
|copying them to an offscreen-GWorld is very slow!
|What is the preffered way for storing and reading the
|graphics without waiting 10-15 seconds (at least on
|my LC475) before the animation starts?

One thing to remember (and this goes for all programs, not just games) is
that it doesn't have to *be* fast.  It just has to *seem* fast.  There are
several ways to increase the apparent speed of your startup (even if you
can't figure out how to increase the actual speed).

1.  Display some nice graphics.  Look at Arashi.  The opening graphics are
so cool that the user doesn't mind watching.  While the graphics are being
displayed---as long as they're not too processor intensive---you can do
your initialization (load sprites & sounds, etc.).

2.  Display a progress bar.  This does a lot to alleviate user
frustration.  The user knows that the time isn't being wasted, that some
necessary startup stuff is being done, and has some idea how long it's
going to take.  (Can you imaging waiting for Maelstrom to start, if it
didn't have a progress bar?)

3.  Use menu bars and dialog boxes.  Many games start up as normal
Macintosh applications.  The user has to select "New Game" from the menu
or from a dialog box to start the game.  Don't waste this time!  You can
be loading graphics while waiting for menu or mouse events.  (If the user
selects "New Game" before you're done loading---well, at least you got
*some* of it done.)

4.  Use between-level screens.  ("Prepare for level 1".)  While the user
is preparing for level 1, you can also be preparing!  This is also a good
time to generate a bunch of random numbers in advance, or make other
necessary calculations.

--
Mark Chrisman
[e-mail]

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

>From [e-mail] (Alex Curylo)
Date: 21 Nov 1995 02:53:01 GMT
Organization: First Ascent

In article <48q72t$[e-mail]>
[e-mail] (Thorsten Goertz) writes:

> The picts are stored in the resource-fork. Reading and
> copying them to an offscreen-GWorld is very slow!

I hear you, brother. This adventure game I'm doing has the neighborhood
of 300 animation frames in some rooms. PICTs and GWorlds start to
become inconvenient in that neighborhood :)

> What is the preffered way for storing and reading the
> graphics without waiting 10-15 seconds (at least on
> my LC475) before the animation starts?

Encoding all the sprites for a room into a big data file as per the
scheme outlined in _Tricks of the Mac Game Programming Gurus_ is
working well enough for me. You read all the data in one big lump, a
nice bonus if you happen to be  playing off CD-ROM, and you just find
the offset of each sprite instead of going through that setting up the
GWorld stuff. Very convenient. Easier on the Memory Manager too ;)

Now, if someone would just post some optimized flipping and scaling
code for that encoder...


Alex Curylo  [e-mail]  [phone], fax -1359
***  First Ascent: Mac programming, UP paragliders, indie CDs  ***

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

>From [e-mail] (Juri Munkki)
Date: 22 Nov 1995 04:37:10 GMT
Organization: Helsinki University of Technology

In article <[e-mail]> [e-mail] (Mark Chrisman) writes:
>1.  Display some nice graphics.  Look at Arashi.  The opening graphics are
>so cool that the user doesn't mind watching.  While the graphics are being
>displayed---as long as they're not too processor intensive---you can do
>your initialization (load sprites & sounds, etc.).

You must be talking about another game... Arashi doesn't really load much
more than the sounds and it doesn't display anything while it's doing that.
By the time the idle animations are running, everything has already been
loaded. Maybe you were thinking of Apeiron or Maelstrom.

--
Juri Munkki [e-mail]      In cyberspace everyone can hear you scream.
http://www.iki.fi/~jmunkki              Windsurfing: Faster than the wind.

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

>From [e-mail] (Ingemar Ragnemalm)
Date: 24 Nov 1995 16:54:44 GMT
Organization: (none)

[e-mail] (Thorsten Goertz) writes:

>I'm currently writing a program that has to display some
>large picts in 256 colors and animates some smaller picts
>as sprites on the screen.
>The picts are stored in the resource-fork. Reading and
>copying them to an offscreen-GWorld is very slow!
>What is the preffered way for storing and reading the
>graphics without waiting 10-15 seconds (at least on
>my LC475) before the animation starts?

Drawing pictures isn't very fast. If you only run in 256 colors and with
a specific color table, you can store the graphics as the pixel maps.

--
- -
Ingemar Ragnemalm, PhD
Image processing, Mac shareware games
E-mail address: [e-mail] or [e-mail]

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

>From [e-mail] (KeeneyBo)
Date: 26 Nov 1995 20:32:42 -0500
Organization: America Online, Inc. (1-800-827-6364)

[e-mail] (Thorsten Goertz) writes:

>I'm currently writing a program that has to display some
>large picts in 256 colors and animates some smaller picts
>as sprites on the screen.
>The picts are stored in the resource-fork. Reading and
>copying them to an offscreen-GWorld is very slow!
>What is the preffered way for storing and reading the
>graphics without waiting 10-15 seconds (at least on
>my LC475) before the animation starts?

Most games read pics and sprites into memory when the game is first
started up while the user is settling down to play the game.  In the many
games I've played over the years this method is common and not too
annoying.  Who cares if I have to wait 30 seconds for the game to
initialize?   As long as you don't continually dispose and reload the pics
I don't think you'll have a problem.

[e-mail]

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

>From [e-mail] (Michael Trent)
Subject: [Q] Color-Picker Popup Menu?
Date: Fri, 17 Nov 1995 00:02:31 -0500
Organization: FullFeed Madison ([e-mail])

Here's a question I've been stewing over for some time:

I want to make a popup menu similar to the color pickers one finds in any
program w/ graphic capabilities. You know what I mean: Go to ResEdit (for
example), Edit a cicn, click on the color rectangles on the left side of
the edit window, and presto! a cool color menu appears from which you can
pick what color you want to plot.

Ok. I understand the MDEF. Not a problem. Actually, it was kind of fun!

The first problem is: I don't want a text title for my popup menu. I want
a framed rectangle containing the color selected from the menu. The second
problem is, I don't know how to do that.

I suppose I could fake it: just draw a rectangle and fill it with the
desired color. But then how would I get the menu to appear? Should the
popup be sitting near-by with a title width of 0? So many people implement
this thing, there has to be a "right way" to do it. Even so, I haven't
found any code examples to help me out.

I'd appreciate a hand in figuring this out. Thanks in advance!

M Trent

--
>>SIG UNDER CONSTRUCTION<<

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

>From [e-mail] (Ram Rajadhyaksha)
Date: Fri, 17 Nov 1995 12:39:12 -0500
Organization: The Ohio State University

In article <[e-mail]>,
[e-mail] (Michael Trent) wrote:

> I suppose I could fake it: just draw a rectangle and fill it with the
> desired color. But then how would I get the menu to appear? Should the
> popup be sitting near-by with a title width of 0? So many people implement
> this thing, there has to be a "right way" to do it. Even so, I haven't
> found any code examples to help me out.

Um, I haven't written an MDEF yet so I have limited knowledge about this
but, couldn't you just call PopUpMenuSelect? Draw the color box yourself
and call that trap when you have a mousedown in the area. It's the "old"
way of doing a pop-up menu and I think it's described in IM V.

I also think Apple distributes a sample source of this...

-Ram

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

>From [e-mail] (Somebody Else)
Date: 18 Nov 1995 00:23:24 GMT
Organization: CYBERCOM Internet Services  [phone]

In article <[e-mail]>,
[e-mail] (Michael Trent) wrote:

> Here's a question I've been stewing over for some time:
>
> I want to make a popup menu similar to the color pickers one finds in any
> program w/ graphic capabilities. You know what I mean: Go to ResEdit (for
> example), Edit a cicn, click on the color rectangles on the left side of
> the edit window, and presto! a cool color menu appears from which you can
> pick what color you want to plot.
>
> Ok. I understand the MDEF. Not a problem. Actually, it was kind of fun!
>
> The first problem is: I don't want a text title for my popup menu. I want
> a framed rectangle containing the color selected from the menu. The second
> problem is, I don't know how to do that.
>
> I suppose I could fake it: just draw a rectangle and fill it with the
> desired color. But then how would I get the menu to appear? Should the
> popup be sitting near-by with a title width of 0? So many people implement
> this thing, there has to be a "right way" to do it. Even so, I haven't
> found any code examples to help me out.

I think what you want is to create a "fake" static menu item (the current
menu selection) , and then have your menu popup, right?  I think at the
moment you're using a popup menu cdef, and you have the attributes set
so that you have the attached label, right? Instead you should forgo the
popup menu control, and do it the "old" way, before the cdef was released.

This is actually a common thing, except for the custom mdef.  What you do is use
PopupMenuSelect().  There's no control or cdef involved.

For the static menu, you draw whatever you want the regular way,
the current color in a box, and a little downward-pointing triangle, or
whatever.
Then when you get a click in the menu's rect, you call PopupMenuSelect()
with your menu , and it operates just like the popup control would.  The trap
will tell you what item the user selected (if any), and then it goes away.   You
then take the new color, and redraw the static menu that way.
--
Uh…okay.

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

>From [e-mail] (Jordan Zimmerman)
Date: Fri, 17 Nov 1995 18:46:13 -0900
Organization: Altura Software, Inc.

In article <[e-mail]>,
[e-mail] (Ram Rajadhyaksha) wrote:

> In article <[e-mail]>,
> [e-mail] (Michael Trent) wrote:
>
> > I suppose I could fake it: just draw a rectangle and fill it with the
> > desired color. But then how would I get the menu to appear? Should the
> > popup be sitting near-by with a title width of 0? So many people implement
> > this thing, there has to be a "right way" to do it. Even so, I haven't
> > found any code examples to help me out.
>
> Um, I haven't written an MDEF yet so I have limited knowledge about this
> but, couldn't you just call PopUpMenuSelect? Draw the color box yourself
> and call that trap when you have a mousedown in the area. It's the "old"
> way of doing a pop-up menu and I think it's described in IM V.
>
> I also think Apple distributes a sample source of this...
>
> -Ram

I've written a color picker MDEF. It's on AOL, alt.mac.sources, etc.

--
Jordan Zimmerman, Altura Software
home page: http://www.altura.com/jordanz
Spock's Beard: http://www.altura.com/spocks_beard

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

>From [e-mail] (Ken Long)
Date: Sat, 18 Nov 1995 08:29:56 GMT
Organization: NETCOM On-line Communication Services (408 261-4700 guest)

Michael Trent ([e-mail]) wrote:

: I want to make a popup menu similar to the color pickers one finds in any
: program w/ graphic capabilities. You know what I mean: Go to ResEdit (for
: example), Edit a cicn, click on the color rectangles on the left side of
: the edit window, and presto! a cool color menu appears from which you can
: pick what color you want to plot.

You've asked about this before, haven't you?

I have worked on a duplicate of that ResEdit doo-dadd, but I keep getting
side tracked, so it's not done.

: The first problem is: I don't want a text title for my popup menu. I want
: a framed rectangle containing the color selected from the menu. The second
: problem is, I don't know how to do that.

There was some old code that made a pie chart, called "PopPieSelect" or
some such (B/W) that could help with that.  My floppy drive is down, so
maybe someone else that has it could send it to you.

: I suppose I could fake it: just draw a rectangle and fill it with the
: desired color. But then how would I get the menu to appear? Should the
: popup be sitting near-by with a title width of 0? So many people implement
: this thing, there has to be a "right way" to do it. Even so, I haven't
: found any code examples to help me out.

You need to get up close and personal with the Tear-Off stuff (straight C
source, buried in the Think Class Lib folder), because that's what the
thing in ResEdit is.

It must work off PtInRect, which then draws the palette off that.  But
then the tear-off stuff takes over if your mouse strays too far while
StillDown.

The DTS.Draw source has a tear-off picture menu (so does Pictoid).

Maybe the rectangle that triggers the palette in ResEdit is a Rgn,
because the background one is partially obscured by the forecround one?

At any rate, whether you use an MDEF or not, the PtInXxxx is what it is
oriented to initially.

That palette is drawn too quick to be a grid of colors, yet there is no
'PICT' resource in ResEdit (that I've seen) representing it.  But a
'PICT' would work.

If you tear it off, you drag a dotted ghost until you let up.  Then the
palette becomes a floater and in the same place.

There must be some pictire or rectangle popup source around, which would
operate similarly, to start off with.

A fixed floater, as in Color It, would be easier, since you wouln't have
to pop it up.

If the source Jordan Zimmerman mentioned is one I have, it is excellent,
and may give great insight as to how to duplicate the ResEdit thing.

Personally, I think Apple should release the source for ResEdit.  What
harm could it do?  The program is free - why not the source?

-Ken-

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

>From Dave Overton <[e-mail]>
Date: Sat, 18 Nov 1995 14:12:31 GMT
Organization: DataStream Imaging Systems, Inc.

Michael,
I'd say the easy way to do it is to use old style PopUps- not the control
system 7 kind.
That way, you can draw your own popup title.
Dave Overton

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

>From [e-mail] (Dave the Rave)
Date: Wed, 22 Nov 1995 16:36:26 GMT
Organization: Unseen University, Ankh-Morpork

In article <[e-mail]> Dave Overton <[e-mail]> writes:
>Michael,
>I'd say the easy way to do it is to use old style PopUps- not the control
>system 7 kind.
>That way, you can draw your own popup title.
>Dave Overton

That's right. The easiest way to do this (the only way as far as I can tell)
in ordinary Toolbox mode is to write your own MDEF, create a menu with
this MDEF ID, and pass the menu to PopupMenuSelect.

There is a set of classes in the Think Class Library that handles floating
windows and tearoffs, and another for handling arbitrary pop-up menus.
Check it out.

______________________________________________________
@-|~ (Scientist formerly known as David A. G. Gillies)
                      (email: [e-mail])
You are in a maze of kinky little fetishes,  all alike
(c) 1995 Wittgenstein's Amazing Underwater Supermarket
_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/_/


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

>From [e-mail] (GrayOwl1)
Subject: [Q] How to hit-test a PICT?
Date: 27 Nov 1995 18:51:03 -0500
Organization: America Online, Inc. (1-800-827-6364)

Does anyone know how to check whether a given point is "inside" a PICT
(the PICT having been obtained from a PICT resource)? By inside, I mean
not just inside the picFrame, but inside the outline of whatever the
picture contains.

I know how to do this with a bitmap--get the mask of the bitmap and test
whether the pixel corresponding to the point is 1--but I can't figure out
how to do this with a PICT, without drawing the PICT into an offscreen
bitmap, converting it to a bitmap, etc.

Any ideas? Please e-mail to [e-mail] as well as this newsgroup.
Thanks in advance,

Regards,

Vinay Prabhakar
Gray Owl Software
- ---------------------------------------------------
"remember my name, I'll be famous someday"- anonymous
- ---------------------------------------------------

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

>From [e-mail] (StarLabs)
Date: 27 Nov 1995 22:37:35 -0500
Organization: America Online, Inc. (1-800-827-6364)

>>
Does anyone know how to check whether a given point is "inside" a PICT
(the PICT having been obtained from a PICT resource)? By inside, I mean
not just inside the picFrame, but inside the outline of whatever the
picture contains.
>>

One method I use is to create a region mask of the pict and save it as
a resource. When running your application, you could then load it when
needed, move it to the appropriate location, and use PtInRgn.

There's the extra effort in creating the region mask beforehand and
saving it as a resource - but I think this is better than drawing the
PICT offscreen and checking the corresponding pixel. It's just a matter
of technique preference in the end. A while back I wrote a utility
just to do what I mentioned: convert a PICT (preferably B&W) to a
region and save it as a resource. Email me if you're interested...


--Hiep

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

>From [e-mail] (JIAN HE)
Subject: multi-player keyboard control
Date: 25 Nov 1995 20:21:10 GMT
Organization: University of California, Berkeley


I want to know to write a multi-player keyboard control game. Right now I
have a maze game for one player. I try to add another player to the game but
there is a problem getting the keyboard events. I am using keyDown
+ autoKeyMask for movement. When one player is holding down a key, the other
one can't move his character. How can I solve that problem?

Jian

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

>From [e-mail] (TheBug)
Date: Mon, 27 Nov 1995 18:54:48 +0100
Organization: privat

In article <497trm$[e-mail]>,
[e-mail] (JIAN HE) wrote:

> I want to know to write a multi-player keyboard control game. Right now I
> have a maze game for one player. I try to add another player to the game but
> there is a problem getting the keyboard events. I am using keyDown
> + autoKeyMask for movement. When one player is holding down a key, the other
> one can't move his character. How can I solve that problem?


The problem you are encountering is that the keyboard does not support
all-key-rollover. That means it can not decode the key matrix if too many
keys are pressed. You should be able to get two typing keys properly plus
all of the modifier keys, everything above that will be erratic. This is
part one of the problem.
Since you are using keyDown+autoKeyMask you don't get the key up events.
So I suppose you are expecting to get repeat events for both keys. The
keyboard driver is only generating repeat events for the key LAST pressed.
I would suggest you use keyDown+keyUp and do the repeat internally with a
timer task.

If you should decide to go to GetKeys instead, then please keep in mind
that there might be more then just the single keyboard generating key
events. If the user has a joystick connected the joystick will also
generate key events. And GetKeys reports only keys of the device that was
the last to send changes.

--
*******************************************************************
Guido Körber -            Programmer and hardware hacker
[e-mail]    Specialised in mistreating the ADB
fax: x49-30-773 81 36     Ask me about:
                           Flightstick Pro, Jetstick
                           MacEnjoy, MacEnjoy Style

Opinions expressed herein are mine unless expressly stated
otherwise. Similarities with living or undead persons are
coincidence and not intended - really!   ;-)
Best use before: (see date printed on backside of message)
*******************************************************************

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

End of C.S.M.P. Digest
**********************